Article Text

Download PDFPDF

Molecular interactions of polo-like kinase 1 in human cancers
  1. Wayne Tiong Weng Ng1,2,3,
  2. Joo-Shik Shin1,2,3,4,5,
  3. Tara Laurine Roberts2,4,
  4. Bin Wang1,2,6,
  5. Cheok Soon Lee1,2,3,4,5,6,7
  1. 1Discipline of Pathology, School of Medicine, Western Sydney University, Sydney, Australia
  2. 2Centre for Oncology Education and Research Translation (CONCERT), Ingham Institute for Applied Medical Research, Sydney, Australia
  3. 3Cancer Pathology and Cell Biology Laboratory, Ingham Institute for Applied Medical Research, Sydney, Australia
  4. 4Molecular Medicine Research Group, School of Medicine, Western Sydney University, Sydney, Australia
  5. 5Department of Anatomical Pathology, Liverpool Hospital, Sydney, Australia
  6. 6South Western Sydney Clinical School, University of New South Wales, Sydney, Australia
  7. 7Cancer Pathology, Bosch Institute, University of Sydney, Sydney, Australia
  1. Correspondence to Wayne Ng, Centre for Oncology Education and Research Translation (CONCERT), Ingham Institute for Applied Medical Research, 1 Campbell St, Liverpool NSW 2170, Australia; W.Ng{at}westernsydney.edu.au; soon.lee{at}westernsydney.edu.au

Abstract

Polo-like kinase 1 (PLK1) is an essential protein in communicating cell-cycle progression and DNA damage. Overexpression of PLK1 has been validated as a marker for poor prognosis in many cancers. PLK1 knockdown decreases the survival of cancer cells. PLK1 is therefore an attractive target for anticancer treatments. Several inhibitors have been developed, and some have been clinically tested to show additive effects with conventional therapies. Upstream regulation of PLK1 involves multiple interactions of proteins such as FoxM1, E2F and p21. Other cancer-related proteins such as pRB and p53 also indirectly influence PLK1 expression. With the high mutation rates of these genes seen in cancers, they may be associated with PLK1 deregulation. This raises the question of whether PLK1 overexpression is a cause or a consequence of oncogenesis. In addition, hypomethylation of the CpG island of the PLK1 promoter region contributes to its upregulation. PLK1 expression can be affected by many factors; thus, it is possible that PLK1 deregulation in each individual patient tumours could be due to different underlying mechanisms.

  • CANCER
  • MOLECULAR PATHOLOGY
  • TUMOUR MARKERS
  • CELL CYCLE REGULATION
View Full Text

Statistics from Altmetric.com

Introduction

The discovery of polo-like kinase (PLK) in 1988 has provided insight into the understanding of the cell cycle and tumourigenesis.1 In humans, five PLKs were discovered (PLK1–5) with different functions in cell cycle and different distribution in tissues.2 Among the PLKs, PLK1 is the most extensively studied. PLK1 is composed of a kinase domain and two polo box domains (PBD). It is a serine/threonine kinase that is critical in cell-cycle processes such as spindle assembly, centrosome maturation, checkpoint recovery, sister chromatid separation, cytokinesis and in particular mitotic entry.2 Upregulation of PLK1 is associated with several types of cancers and indeed PLK1 had been validated as a prognostic marker in some of these cancers.3 The lack of mutation that causes the increased activity of PLK1 created arguments on whether the overexpression of PLK1 seen in some cancers is a cause or a consequence of oncogenesis.4 ,5 Nevertheless, PLK1 is still an important target for cancer pathology and anticancer treatment. The molecular interactions of PLK1 have been studied extensively throughout recent years. This article aims to review and update the roles of PLK1 in cell-cycle regulation, DNA damage responses, oncogenesis and also highlights the interactions that occur in the promoter region of PLK1. Finally, we will discuss the latest findings on PLK1 inhibitors.

PLK1 in cell cycle and cell growth

In proliferative cells, PLK1 expression is initiated in late S phase and accumulates through G2 and M phases of the cell cycle. The kinase activity of PLK1 reaches maximum during M phase. PLK1 plays a critical role in the progression and timing of mitosis by influencing the CDK1–cyclin B complex activity;6 ,7 the latter is important in controlling cell entry into M phase. Even though the activity of the complex relies heavily on the presence of cyclin B during M phase entry, phosphorylated CDK1 prevents the complex from promoting M phase. While CDC25 phosphatase dephosphorylates CDK1 to activate the complex, it is counteracted by WEE1 and MYT1. Although CDK1 activation provides a positive feedback loop to activate CDC25, PLK1 is the key protein in activating CDC25 and inhibiting WEE1 and MYT1, therefore promoting the activation of CDK1. These molecular interactions are shown to have similar patterns involving other orthologues of PLK1 such as Plo1,8 Cdc59 and Plx1,10 indicating the conservation of PLK1 roles. In fission yeast, Cut12 proteins facilitate the recruitment of Plo1 to spindle pole bodies (centrosome of yeast) to allow successful activation.11 ,12 These reactions involve mammalian target of rapamycin signalling via protein kinase B (Akt) or ribosomal protein S6 kinase β-1 (S6K) in which timing for mitotic entry is determined, depending on the cell conditions (eg, stress or normal). Aurora A kinase, which is activated by protein aurora borealis (Bora) serves as an activator for PLK1 in cell cycle.13 The schematic summary of the molecular interactions of PLK1 is shown in figure 1A.

Figure 1

Summary of molecular interactions of PLK1 (A). Green arrows indicate the activations, whereas red arrows indicate the inhibitory effects. Promoter region of PLK1 gene and the critical binding sites for activation or suppression of PLK1 expression (B).

PLK1 in DNA damage response

DNA damage can be induced by endogenous and exogenous factors such as replication stress and ultraviolet irradiation. If the DNA damage response is triggered during S/G2 interphase, cell division is halted by G2 DNA damage checkpoint and PLK1 remains inactive. Activated ATM/ATR phosphorylates Bora at threonine 501 residue, resulting in degradation of Bora by E3 ubiquitin ligase SCF-b-TRCP.14 PLK1 cannot be phosphorylated by Aurora A without the facilitation of Bora, leading to inhibition of Cdc25 phosphatase, which in turn deactivates CDK1 that signals the mitotic entry. Other than that, ATM/ATR pathway also inhibits Cdc25 phosphatase via Chk1 and Chk2. Once DNA damage is repaired, PLK1 activity is restored to negatively regulate the ATR pathway by dissociating Claspin from ATR, hence allowing for resumption of mitotic entry.15 However, there are conflicting reports on whether PLK1 inactivation is dependent on the ATM/ATR pathway.16–18

DNA damage also triggers the stabilisation and activation of the tumour suppressor p53 via the ATM/ATR and Chk1/Chk2 pathways.19–21 p53 plays a critical role in signalling cell-cycle arrest and apoptosis during the DNA damage response. PLK1 and p53 functions are counteractive. Murine double minute 2 (Mdm2) suppresses p53 by proteasomal degradation.22 G2 and S phase expressed protein 1 (GTSE1), which is known to link with Mdm2 for the regulation of p53, was found to be phosphorylated by PLK1 at Ser435, leading to its translocation to nucleus.23 GTSE1 then binds to and shuttles out p53 from the nucleus, preventing its induction of cell-cycle arrest and apoptosis. In addition, PLK1 also downregulates p53 through phosphorylation of Topors at Ser718.24 The phosphorylated Topors decrease in small ubiquitin modifier (SUMO) E3 ligase activity but rise in ubiquitin activity, therefore increasing the ubiquitination and degradation of p53.

On the other hand, p53 regulates the transcription of PLK1 as well as other mitotic regulators such as cyclin A and cyclin B. Stabilised p53 is localised to the promoter region of PLK1, binding to E2F and thus preventing E2F from promoting the transcription of PLK1 by forming p53–E2F–PLK1 promoter complex.14 In addition, p53 also activates p21/Waf1 that binds to the promoter region of PLK1 and inhibits PLK1 at the transcriptional level. The interactions between p53, p21 and the promoter region of PLK1 are discussed later. This negative regulation can be counteracted by forkhead box protein (FoxM1) when DNA damage is resolved. FoxM1 is activated through phosphorylation at Ser715 and Ser724 by PLK1.25 Therefore, PLK1 is suggested to be important in recovering from DNA damage.

The DNA damage response in mitosis is different from interphase.26 DNA damage does not initiate cell-cycle arrest but instead triggers a delay in cell division to allow ample time for DNA repair.27 ,28 Similar to the response in interphase, ATM and ATR are still activated by DNA damage in mitotic cells, and yHA2X, a phosphorylated protein which binds to sites of DNA double-strand breaks, remains located in foci at the breaks even though the cells are dividing.29 PLK1 activity is also reduced via the activated ATM/ATR pathways.30 However, since PLK1 is in active mode throughout mitosis, it appears to be deactivated by protein phosphatase 2A.31 PLK1 is believed to be involved in communicating the DNA damage with mitotic checkpoints.26 The reduction of PLK1 activity was suggested to prolong mitosis, therefore allowing DNA to be repaired. If the DNA damage is too severe, the cell would bypass late mitosis and cytokinesis and enter G2-like G1 phase with tetraploid (4N) DNA contents and positive y-H2AX signal.32 The DNA contents are then replicated to 8N. Subsequently, the cell would be signalled to undergo apoptosis after a cycle of replication. PLK1 is also suggested to suppress DNA damage response during mitosis by inhibiting 53BP1 that is required for homologous recombination (HR) and non-homologous end joining DNA repair.33 53BP1 was found phosphorylated by PLK1 when DNA damage was induced in mitosis.34 Cdk1 phosphorylates PLK1 to create a docking site for the phosphorylation of 53BP1.

An interesting discovery about PLK1 is its role in DNA repair. The recruitment of radiation-sensitive 51 homologue 1 (Rad51) to the DNA damage site for the Mre11 trimer formation may require the kinase activity of PLK1.35 Rad51 is phosphorylated by PLK1 at Ser14 and subsequently phosphorylated by casein kinase (CSKN2) at Thr15. The dual phosphorylation enables Rad51 to be recruited to the damage site and thereby initiate HR repair. In conjunction with other findings related to the involvement of PLK1 in cell recovery, the findings suggest that PLK1 is not completely shut down during DNA damage response. A minimal PLK1 activity is maintained to initiate DNA repair and recovery control.

Promoter region of PLK1

Experiments have shown that PLK1 expression and activity during the cell cycle is regulated at the transcriptional level.36 ,37 Deregulation of PLK1 transcription is believed to be the major cause of PLK1 overexpression in cancers. Understanding the promoter region of PLK1 and the associated interacting proteins will provide insight into PLK1 mRNA regulation (figure 1B). During cell cycling, PLK1 mRNA accumulates at S phase, peaks in M phase and drops after mitosis. The variation in expression is sixfold to eightfold between the peak and the lowest levels.36 The untranslated region upstream of the transcription initiation site consists of 2.3 kb with enhancer, silencer and promoter regions.38 Core portion within the promoter region mediates up to 70% of the promoter activity. Mutation and deletion studies of the promoter region showed that sequences between -93 and +65 (+1 as transcription initiation site) are critical for the induction of PLK1 mRNA in G2/M phase, whereas the nucleotides from −93 to −35 are responsible for the minimal promoter activity of PLK1 in G1 phase.38–40 Three elements including a non-consensus site, GC-rich stretch and a CCAAT containing site were identified within this region (figure 1B). The CCAAT site is required for p21-dependent PLK1 suppression via binding of nuclear factor Y (NF-Y). NF-Y is also found in other cell-cycle-related genes such as CDK1, E2F1, CDC25C and CCNA2.40 The suppressing mechanisms are discussed later. Moreover, further upstream of the promoter region is shown to have transcription factor FoxM1 binding sites,41 which are also responsible for increasing PLK1 mRNA in G2-specific manner.42 In addition, the promoter region also contains a binding site for transcription factor E2F that is believed to initiate PLK1 mRNA expression.36 ,43

The transcriptional suppression of PLK1 is critical for maintaining cell-cycle integrity. The key negative regulating elements of PLK1 promoter are located right on the transcription initiation site and are called cell-cycle-dependent element and cell-cycle gene homology region (CDE/CHR, figure 1B).36 CDE/CHR element is found in several G2/M-specific genes like cyclin B1, cyclin A1, Cdc2, Aurora A and Cdc25C. Mutational studies showed that the sequence is important for cell-cycle-dependent repression of PLK1 gene.39 The transcriptional repression of PLK1 is known to occur through binding of p21 on the CDE/CHR element. Furthermore, p53 that mediates the expression of PLK1 becomes an indirect regulator of PLK1 gene. NF-YA binds to CCAAT box to act as activator or repressor for the targeted genes.44 Recent findings suggest that the suppression is due to the formation of p21 and NF-YA subunit axis. In stressful conditions, p21 binds to CDK2 sitting on the CCAAT site, such that NF-YA can bind to the site and form p21–NF-YA–PLK1 promoter axis, leading to suppression of PLK1.45

PLK1 and oncogenesis

Since PLK1 plays major roles in cell cycle, maintaining genome stability, communicating between DNA damage and recovery competence, it is not surprising that deregulation of PLK1 expression is associated with carcinogenesis.

In contrast to normal human cells in which PLK1 is only expressed in G2/M phase, PLK1 is also detected in G1/S phase in cancer cells. In addition, PLK1 was shown to be involved in DNA replication during S phase.46 ,47 Since mutation of PLK1 is rare in cancers (0.47% of samples deposited to COSMIC)48 and there is a lack of correlation between mutations with the elevation of PLK1, the abnormal regulation of PLK1 could be due to interactions of its upstream proteins or DNA methylation of CpG island that affects the promoter activity of PLK1.

Cdks and their endogenous inhibitors are critical in regulating cell-cycle phases to ensure proper cell division. They appear upstream and downstream of PLK1 activity, suggesting the interactions may affect PLK1 expression (figure 1A). Cdk4/6 and cyclin D1 are activated in G1 phase, which in turn inhibit Rb protein,49 leading to activation of E2F and therefore increasing the transcription of cyclin E, cyclin A, DNA polymerase and PLK1.50 Cyclin D1 has been reported as a proto-oncogene, with upregulated cyclin D1 detected in many cancers.51 Rb protein is also a major tumour suppressor protein that is mutated in various cancers.52 Mechanistically, Cdk4/6-cyclinD1 complex binds to Rb protein, resulting in dissociation of Rb from E2F, releasing free activated E2F for initiation of transcription of cell-cycle genes.49 Inhibitor of Cdk4/6 (INK4 family) and Cdk-interacting protein/kinase inhibitor protein (KIP) families are the upstream regulators of Cdks; therefore, they may play a role in indirectly or directly regulating PLK1. The aforementioned p21 protein is one of the KIPs, and since p21 is downstream of p53 protein activity, PLK1 and Cdks are affected by p53 activity. Mutations of INK4 member p16 are detected in various cancers, resulting in loss of p16.4 Cdk4/6 is deregulated due to loss of p16 activity, leading to indirect activation of E2F and then PLK1.

The methylation status of CpG islands of the promoter in many genes determines their expression levels. Deregulation of PLK1 expression has also been associated with hypomethylation at the promoter region of PLK1 gene. Hypomethylation of PLK1 promoter has been observed in mouse hepatocellular carcinoma53 ,54 and human haematological malignancies.55 The loss of methylation correlated (75.7% in haematological malignancies) with elevated PLK1 expression in tumour cells. In addition, oxidative stress induces methylation of the CpG island of PLK1 resulting in decreased PLK1 expression. Nevertheless, in human liver the methylation status of normal and tumour tissues affects PLK1 transcription. Considering the negligible rate of mutation and the correlation between DNA methylation and variable PLK1 expressions in cancer cells, epigenetics may play a role in deregulating PLK1 at the transcriptional level. Nonetheless, the up-to-date findings related to PLK1 methylation are still preliminary and conflicting.

Loss of PLK1 is also associated with oncogenesis.56 Missense mutations at PBD,5 which functions as the mediator for kinase activity of PLK1 in some cell lines, hinder the molecular folding of PLK1 by heat shock protein 90, resulting in unstable PLK1 and therefore low expression of PLK1.57 Tumour formation is more likely in mice with heterozygous PLK1 gene, suggesting that balanced PLK1 expression is required to avoid tumourigenesis.58 While overexpression of PLK1 is linked to uncontrolled cell proliferation, low expression of PLK1 results in improper cell-cycle processes such as spindle assembly, centrosome maturation and so on, both leading to tumour progression.

PLK1 in human cancers

Overexpression of PLK1 is observed in tumours of different origins (reviewed in ref. 3). Tumours with higher expression of PLK1 are often linked with poorer prognosis and lower overall survival.

Colorectal cancer

Overexpression of PLK1 had been observed in patients with colorectal cancer in multiple studies.59–62 Increased expression of PLK1 is associated with poorer prognosis and therefore reduced patient survival. In our group, we found inferior survival outcomes in patients with rectal cancers displaying high PLK1 expression.62 Furthermore, in vitro experiments show that RNAi knockdown of PLK1 reduces radio-resistance of colorectal cancer cells, implying that the poorer survival outcome in patients with increased PLK1 expression is related to increased radioresistance of the tumour cells (unpublished observations). Similarly, Rodel et al61 also reported that higher PLK1 expression in rectal cancer is associated with poorer radiotherapy response, and that RNAi knockdown of PLK1 radiosensitises colorectal cancer cells.

Lung cancer

In an earlier study of non-small cell lung cancer, the 5-year survival rate of the patients with tumours with moderate expression of PLK1 was higher (51.8%) than those with a high level of PLK1 transcript (24.2%) in the tumours.63 An in vitro PLK1 knockdown study using recombinant plasmid containing antisense RNA targeting PLK1 also showed that knockdown of PLK1 in A549 cells resulted in growth inhibition, G2/M arrest and apoptosis.64

Breast cancer

Prognosis of breast cancer has also been associated with PLK1 expression. Patients with high PLK1 expression in breast tumours have lower survival rate compared with those with low PLK1 expression.65 In addition, PLK1 expression is associated with TP53 mutation and patients with high PLK1 expression and TP53 mutation have the worst survival.65 ,66 Furthermore, BRCA2 is phosphorylated by PLK1, leading to change of function to mitotic progression instead of DNA repair,67 which is important for maintaining DNA integrity of the cells. Malfunctioning BRCA2 is associated with poorer prognosis in breast cancer.

Ovarian cancer

Similarly, PLK1 expression has been shown to be associated with prognosis of ovarian cancer.68–70 As in breast cancer, PLK1, p21 and p53 were shown to correlate with the prognosis of epithelial ovarian cancer. Patients with high PLK1 and high p53 or high PLK1 and low p21 have the lowest survival rate.70

Other cancers

High PLK1 expression in other cancers such as endometrial cancer71 and gliomas72 is also associated with poorer prognosis. Tategu et al43 reported downregulation of PLK1 in most adult tissues except testis, liver and thymus, whereas most cancer cell lines from different tissue origins have elevated PLK1 levels.

PLK1 inhibitors: role in chemoradiotherapy

Since PLK1 overexpression correlates with poor prognosis in various cancers, finding inhibitors to target PLK1 has become a major area of research. Several compounds have been suggested to have inhibitory effects on PLK1. The majority of these compounds target the kinase activity of PLK1 (reviewed in ref. 73–76). Some inhibitors such as scytonemin, wortmannin, LY294002, morin and quercetin are multiple kinase inhibitors that inhibit PLK1 as well as other proteins.77–79 The following are some of the small molecule inhibitors that have been proposed to more specifically target the kinase activity of PLK1.

BI-2536

This dihydropteridinone derivative is the first PLK1-specific inhibitor discovered with an IC50 of 0.8 nM in inhibiting enzymatic activity of PLK1, and with more than 10 000-fold selectivity for PLK1 against other kinases. In preclinical studies, this ATP-competitive kinase inhibitor reduced cell proliferation of several cancer cell lines and strongly suppressed the tumour growth in multiple human cancer xenograft models.80 ,81 At the molecular level, the inhibitor acts by causing the cell to form a monopolar spindle, leading to mitotic arrest and eventually apoptosis. Phase I clinical trials of the inhibitor show moderate efficacies in different advanced cancer patients.82 However, clinical data from phase II trials indicates poor efficacy of BI2536 in various advanced cancers (eg, only 2.3% had partial response in unresectable exocrine adenocarcinoma of the pancreas).83 The monotreatment trial was then terminated due to issues with efficacy, safety and pharmacokinetics. Subsequently, an improved inhibitor Volasertib (BI 6727) has been developed.

Volasertib

Similar to BI-2536, Volasertib is a dihydropteridinone derivative that exerts similar antitumour effects in vitro and in vivo.84 Monotreatment phase I/II trials showed limited partial response and stable disease with reduced adverse effects (summary from ref. 73). Subsequent phase I and II trials with combination treatments (such as cisplatin and cytarabine) show improvements in efficacies of response and overall survival for those who responded.85 Many of the clinical trials for Volasertib are still being actively conducted.

Rigosertib (ON 01910.Na)

Preclinical work showed that the molecule is a multikinase inhibitor.86 It inhibits PLK1 as well as the phosphoinositide 3 kinase/AKT pathway. Multiple phase I/II trials show positive outcomes with partial responses in a few patients and stable disease in several patients (summary from ref. 73). Phase III trial is currently ongoing in patients with myelodysplastic syndrome.

GSK461364

The imidazotriazine derivative is an ATP-competitive PLK1 inhibitor and has been shown to exert transient antitumour effect in an in vivo tumour model.87 The tumour growth resumes when the treatment is ceased. Interestingly, cell lines with p53 mutations are more sensitive to the compound, suggesting that the compound may potentially target p53-related tumours.88

Other molecules including TAK-960, MK-1496, NMS-P937 and TKM-080301 were also discovered in preclinical studies.73 The outcomes of clinical trials for TAK-960 (NCT01179399) and MK-1496 are disappointing in terms of efficacy and associated severe adverse effects. TKM-080301 is a PLK1 targeting drug containing siRNA and it is now in phase I/II trials. NMS-P937 has just been put onto a phase I trial. Other compounds such as Poloxin89 and Purpurogallin90 inhibit PLK1 by targeting PBD of PLK1 in preclinical data. The development is, however, still in its infancy. In conclusion, targeting PLK1 as an option to treat cancer seems promising, especially when it is combined with conventional treatment. However, predictive markers of treatment response and the associated side effects will need to be determined to improve the outcomes.

Conclusion

In recent years, the majority of the work on PLK1 focused on developing inhibitors to target PLK1 as a form of cancer therapy. Some groups including ours are interested in validating PLK1 as a predictive marker for radiosensitivity in cancers, since PLK1 has critical roles in DNA damage recovery. Nonetheless, whether elevated PLK1 is a cause or a consequence of oncogenesis remains unclear. Since no mutation is associated with upregulation of PLK1, the upstream protein interactions that interact with the promoter of PLK1 are the potential causes for deregulation of PLK1 expression. However, a number of these upstream proteins are broad effectors for cancer-related pathways, such as p53 and pRb. In addition, epigenetic changes in PLK gene may also contribute to the deregulation of PLK1, making it challenging to characterise the underlying molecular mechanism, and elucidation of these changes may give new insight into the roles of PLK1 in tumourigenesis, and perhaps into targeting PLK1 for cancer therapy.

References

View Abstract

Footnotes

  • Handling editor Runjan Chetty

  • Contributors WN drafted, reviewed and edited the article. J-SS, TLR, BW and CSL reviewed and edited the article.

  • Funding WN is a recipient of a PhD scholarship funded by Cancer Institute NSW. The Centre for Oncology Education and Research Translation (CONCERT) is a translational research centre wholly funded by Cancer Institute NSW.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.