Multifaceted Polo-Like Kinases: Drug Targets and Antitargets for Cancer Therapy

REVIEWS

Multifaceted polo-like kinases: drug targets and antitargets for cancer therapy

Klaus Strebhardt Abstract | The polo-like kinase 1 (PLK1) acts in concert with cyclin-dependent kinase 1–cyclin B1 and Aurora kinases to orchestrate a wide range of critical cell cycle events. Because PLK1 has been preclinically validated as a cancer target, small-molecule inhibitors of PLK1 have become attractive candidates for anticancer drug development. Although the roles of the closely related PLK2, PLK3 and PLK4 in cancer are less well understood, there is evidence showing that PLK2 and PLK3 act as tumour suppressors through their functions in the p53 signalling network, which guards the cell against various stress signals. In this article, recent insights into the biology of PLKs will be reviewed, with an emphasis on their role in malignant transformation, and progress in the development of small-molecule PLK1 inhibitors will be examined.

More than two decades ago polo, the founding member Early observations on the overexpression of PLK1 of the family of polo-like kinases (PLKs), was identified in human tumours12 and on the inhibition of cellu- as having an essential role in the ordered execution of lar proliferation by microinjecting antibodies against mitotic events in Drosophila melanogaster 1,2. Since the PLK1 into HeLa cells13 initiated a series of follow-on discovery of polo, a wealth of functional information on studies using a broad spectrum of inhibitors (for exam- PLKs has been collected in a wide phylogenetic space. ple, dominant-negative forms of PLK1, antisense oli- Five mammalian PLK family members have been iden- gonucleotides and small interfering RNAs) aiming to tified so far, PLK1 (also known as STPK13), PLK2 (also evaluate PLK1 as a potential target for the treatment known as SNK), PLK3 (also known as CNK, FNK and of cancer14–17. The identification of PLK1 as a crucial, PRK), PLK4 (also known as SAK and STK18) and PLK5 functional node in the oncogenic network, in which its (Refs 3–5). An early comparison of murine PLK1 with inhibition results in systemic failure (that is, blockage of POLO and PLK2 uncovered two conserved domains in the tumorigenic state by apoptosis)18 proved to be key PLK1, the canonical serine/threonine kinase domain in developing successful therapies. Subsequently, the and a non-catalytic region, which was later termed the pharmaceutical industry has initiated intensive efforts polo box domain (PBD)6. All PLKs exhibit the same to develop potent and specific small-molecule PLK1 domain topology with the kinase domain located at inhibitors. This development is fuelled by previous the amino-terminal followed by the carboxy-terminal pharmaceutical and clinical investigations, which had PBD (BOX 1). Recent advances in the basic biology of convincingly shown that antimitotic drugs are success- PLK1 have greatly improved our understanding of its ful chemotherapeutic tools for the fight against cancer19. regulation, targets and functions. As reflected by multi- PLK1 inhibitors have been clinically assessed and hold

Department of Obstetrics ple localizations PLK1 has been shown to have a pivotal great promise for the improved treatment of patients 18,20,21 and Gynaecology, School role in regulating the cell cycle, including entry into with cancer . of Medicine, J.W. Goethe mitosis, centrosome maturation, assembly of the bipo- Although PLK1 inhibitors have already reached University, Theodor Stern lar spindle, sister chromatid splitting, activation of the the clinic, analysis of their specificity indicates that Kai 7, 60590 Frankfurt, anaphase-promoting complex/cyclosome (APC/C) the related family members PLK2 and PLK3 are in the Germany. e-mail: strebhardt@em. and exit from mitosis with the initiation of cytokine- range of their inhibitory potential. Despite tremen- uni-frankfurt.de sis (fIG. 1). These aspects have been the subject of many dous progress in understanding the biology of PLK1, doi:10.1038/nrd3184 excellent reviews3,4,7–11. the functional roles of other PLK family members,

Box 1 | Structural insights into the polo-like kinase (PLK) family proteins

All members of the PLK family a b contain an amino-terminal kinase domain and a carboxy-terminal polo box domain (PBD), which is unique to the PLK family. The crystal Ser137 Val210 structure of the kinase domain of human PLK1, in which the regulatory phosphorylation site Thr210 was mutated with the aim of reducing conformational heterogeneity, has been determined in complexes with the non-hydrolysable c ATP analogue adenylylimi- dodiphosphate, the pyrrolo-pyrazole inhibitor PHA 680626 (a) or the ATP-competitive BI 2536 (Refs 157,159). In another approach, the wild-type kinase domain of PLK1 was co-crystallized with the designed ankyrin-repeat protein 3H10 that is selective for PLK1 (Ref. 218). The similarity of the crystal structures confirms the validity of the different approaches for structure-guided drug design. Three-dimensional analysis showed the typical structure of a kinase fold, in which the ATP-binding site is situated in a cleft, formed between the N-terminal lobe encompassing β-sheets (residues 37–131)Na andture theRevie predominatelyws | Drug Discov ery α-helical C-terminal lobe (residues 138–330)157. The two lobes are linked by a hinge region that interacts with the bound ligands. Although Thr210 is not phosphorylated, the activation loop of PLK1 (T-loop; residues 194–221) exhibits an extended conformation in the crystal structure reminiscent to the conformation of activated kinases219. This conformation might be stabilized by Zn2+ ions, which are essential for the crystallization of the PLK1 kinase domain157. The C-terminal PBD functions as a phosphopeptide-binding module by targeting serine-(phosphothreonine/ phosphoserine)-(proline/X) motifs34,41. PLK1, PLK2, PLK3 and PLK5 have a PBD that consists of two polo boxes, whereas PLK4 has only one polo box. The crystal structures of the PLK1 PBD, complexed with an optimal phosphoserine-containing peptide or a peptide corresponding to the binding region of PLK1 within the biological interaction partner CDC25, document an interaction along a positively charged cleft, which is formed between the two polo boxes (b)41,220,221. Each conserved polo box consists of a six stranded β-sheet and an α-helix, which associate to form a 12-stranded β-sandwich domain (c). Recent studies on the crystal structure and binding properties of the PLK1 PBD in complex with the minimal phosphopeptides suggest that the C-terminal serine-phosphothreonine dipeptide provides the high-affinity anchor, whereas the N-terminal residues contribute to the specificity and affinity of the interaction222. The crystal structure of the PLK4 PBD containing a single polo box suggests that this domain can dimerize223. Further analysis has corroborated this model, showing that PLK4 forms homodimers in vivo223. Despite its homodimerization PLK4 remains capable of localizing to the centrosomes and the cleavage furrow during cytokinesis223. Part a of the figure shows the crystal structure of the PLK1 kinase domain with the pyrrolo-pyrazole inhibitor PHA 680626 at 2.4 Å resolution. The N-terminal and C-terminal extensions are coloured orange and the activation loop is coloured green157. The positions of the mutated Val210 (corresponding to the main phosphorylation site Thr210 in the activation domain) and Ser137 (which has been proposed as an additional phosphorylation site for the activation of the kinase activity of PLK1) are indicated. Part b of the figure shows the crystal structure of the PLK1 PBD in a ribbon presentation, in complex with phosphothreonine- containing peptide that is shown as a ball and stick representation in yellow41. Polo box 1 and polo box 2 are shown in red and purple, respectively, and from two different angles. The polo cap at the N-terminal end of polo box 1, which is indicated in grey, folds around polo box 2 and therefore tethers it to polo box 1. The two polo boxes form a pocket to accommodate the 41 phosphopeptide. Part c of the figure is a superposition of the polo box 1 and polo box 2 β6α structures, coloured as in part b .

especially of PLK2 and PLK3, remain elusive. However, important to carefully analyse the roles of PLKs in the in cancer cells PLK2 and PLK3 emerge as important network of stress signalling in cancer cells. mediators of the stress phenotypes in response to DNA After summarizing the relevant biological back- damage or oxidative stress. increasing evidence sup- ground of PLKs and their roles in cancer cells, this ports a mutual crosstalk between all members of the Review will discuss the status quo of small-molecule PLK family and the tumour suppressor p53 under development targeting the kinase domain and the PBD stressful conditions in cancer cells22–24. As the classical of PLK1. Furthermore, this article will consider the clinical hallmarks of cancer are well documented25,26, it will be development of novel PLK inhibitors.

G2 phaseProphasePrometaphase Metaphase AnaphaseCytokinesis

Mitotic entry Furrow ingression and spindle elongation Centrosome maturation, microtubule nucleation and kinetochore capture Figure 1 | Localization and selected functions of PLK1 during mitosis. Indirect immunofluorescence of polo-like kinase 1 (PLK1) localization (green) in HeLa cells during the main phases of mitosis (upper panels). The lower panels depict the additional staining of DNA (blue), α-tubulin (red) and kinetochores (stained with centromere-specificNature Reviews autoantibodies, | Drug Discovery purple).

The role of PLK1 in cellular signalling networks and phosphopeptide mapping have shown that the in human tissues, PLK1 is expressed in dividing cells12,27. phosphorylation of Thr210 (Thr201 in the Xenopus Mice with a homozygous deletion of Plk1 are not viable, homologue Plx1) in the activation loop (T-loop) of the and mice heterozygous for Plk1 develop tumours (BOX 2), kinase domain is required for the activation of PLK1 thereby supporting the vital function of PLK1 for the (Refs 43–45). Several kinases, including protein kinase proliferation of normal cells. An overview of the known A, the Ste20-like kinase and xPlkk1, have been consid- cellular functions of PLK1 is provided in BOX 3. ered activators of PLK1 and its Xenopus homologue The cellular interaction partners and the extent of Plx1, but the evidence for this assumption originates the PLK1 interactome were determined in proteomic only from in vitro data44,46,47. impressive studies describ- screens using the PLK1 PBD as an affinity capture ing the crosstalk among PLK1, the kinase Aurora A and agent; mass spectrometry identified 622 proteins that its binding partner bora have considerably improved exhibited phosphorylation-dependent, mitosis-specific our understanding of the initial activation of PLK1 interactions with PLK1 (Ref. 28). it has been shown that (Refs 48–50) (fIG. 2). Although bora does not significantly alterations in the PBD disrupt the subcellular localiza- change the activity of Aurora A towards histone H1, it tion of PLKs and affect their functions. For example, dramatically enhances the phosphorylation of PLK1 at alterations can upset spindle formation or activation Thr210 by Aurora A49,50 (fIG. 2). This result suggests that of the APC/C by regulating the stability of the APC/C bora induces a conformation of PLK1 that is favourable inhibitor protein EMi1 (also known as FBX05) (fIG. 1, for the phosphorylation by Aurora A, probably by reliev- BOX 3), because PBD-mutated PLKs lose their ability to ing the inhibitory impact of the PBD. However, an exact dock at specific scaffolding proteins and substrates29–35. model for this activation would require the analysis of a Pro-directed kinases like cyclin-dependent kinase 1 crystal structure of the binary complex. (CDK1)–cyclin B1, which phosphorylate serine/threo- Concomitant with the high activity of PLK1 during nine residues preceding a proline residue, generate a mitosis, the transcription factor forkhead box M1 phosphorylated docking site for PLK1 in a process called (FOXM1, which controls the execution of a mitotic non-self-priming. Hence, CDK1–cyclin B1 has a pivotal programme by regulating the expression of a cluster role in directing PLK1 PBD to its subcellular locations. of g2/M target genes) is phosphorylated in its trans- Notably, recent data describe a self-priming process in activation domain by PLK1. The phosphorylation of the late stages of mitosis when CDK1–cyclin B1 activity FOXM1 is preceded by the initial priming phosphor- fades, whereby PLK1 creates its own docking site by ylation of CDK1 (Refs 51–54). Foxm1 knockout or dis- phosphorylating the proteins mitotic kinesin-like pro- ruption of FOXM1 phosphorylation by CDK1/CDK2 tein 2 (MKLP2; also known as KiF20A), PBD-binding induces multiple cell cycle defects, eventually leading to protein 1 (PBiP1) and microtubule-associated protein the formation of aneuploid cells51,53. Thus, the elevated regulator of cytokinesis 1 (PRC1)36–38. PLK1 activity often observed in cancer cells could lead in addition to the spatial regulation29,39, early reports to altered transcriptional programmes, thereby sup- also indicated that the PLK1 C-terminal domain con- porting increased proliferation and compromising taining the PBD can negatively regulate the enzymatic genomic integrity. PLK1 has also been shown to inter- activity of PLK1 (Refs 39–41), which rises at the g2/M act with the tumour suppressor p53 and data describ- transition and reaches a maximum level during mito- ing the interplay of PLK1 and p53 is becoming ever sis42. Mutational studies, mass spectrometric analyses more complex55,56.

Box 2 | Modelling the functions of polo-like kinases (PLKs) in mice PLK1 Plk1 heterozygous mice do not yield Knockout mice Defective No division beyond homozygous null mice, indicating that Plk1 –/– embryonic homozygous Plk1 knockout mice are not 4/8 cell stage development, viable224. Further analysis showed that Plk2–/– dwarfism Plk1–/– embryos do not develop beyond Plk3–/– Increased weight, the eight cell stage, showing that PLK1 tumour development is essential for the cycling of primary at advanced age Plk4–/– Arrest after mammalian cells224. This is similar to the gastrulation at E7.5 essential role of the PLK1 counterparts 1,225,226 Healthy and in Drosophila and in yeasts . Haploinsufficient mice fertile, tumour The embryonic cells were also unable Plk1+/– development to assemble a normal spindle apparatus. at advanced age +/– Intriguingly, 27.5% of the Plk1 mice Plk2+/– developed tumours, most frequently Viable these were lung-invading and +/– Plk3 Viable liver-invading lymphomas. Analysis of chromosome spreads of spleen-derived Plk4+/– Tumour development cells from 6-month-old mice showed at advanced age aneuploidy, supporting the role of PLK1 E3.5 E10.5 P0 Adult in genomic stability. PLK2 Nature Reviews | Drug Discovery The initial characterization of PLK2 in adult mouse tissues identified low transcript levels in the brain, the lung and the heart97. In a comprehensive survey, human PLK2 transcripts were identified, with the highest frequency in the fetal lung, kidney, spleen and heart227. They were also identified in the adult spleen, testis, mammary gland and brain. A study of Plk2–/– mice showed that PLK2 is not essential for embryonic viability and postnatal growth102. However, a role in embryonic development has been proposed as embryos displayed a delay in skeletal development and retarded growth. This was attributed to an impaired trophic function in the placenta102. Plk2–/– mouse embryonic fibroblasts proliferated slowly and displayed a delayed entry into S phase. It is possible that PLK3 compensated for PLK2 in Plk2-deficient mice, a hypothesis that could be tested in Plk2/Plk3 double knockout mice. PLK3 Although the examination of 16 human tissues showed a moderate level of PLK3 mRNA in the placenta and low levels in the ovary, the lung and in peripheral blood leucocytes, most tissues were devoid of PLK3 transcripts115. In newborn mice, low levels of Plk3 transcripts were identified in the heart and elevated levels were identified in the intestine, kidney, liver, lung and skin114. In adult murine tissues, Plk3 mRNA was expressed strongly in the skin and at low levels in the ovary, lung, intestine and brain. Ageing Plk3–/– mice displayed increased weight. Most interestingly, they also developed tumours with the highest incidence in the lung but also in the kidney, the liver and the uterus132. PLK4 In embryonic and adult tissues, Sak-a transcripts coding for PLK4 were observed exclusively in proliferating cells, consistent with the expression profile of Plk1 mRNA137,228. The development of Plk4–/– embryos arrested after gastrulation at embryonic day 7.5 (E7.5) and there was a marked increase in the number of mitotic and apoptotic cells in the embryo139. Phosphorylated histone H3 and cyclin B1 were seen more frequently on the chromatin in late mitotic cells in sections of Plk4–/– embryos than in sections of wild-type embryos, suggesting a role for PLK4 in the anaphase-promoting complex/ cyclosome (APC/C)-dependent breakdown of cyclin B1 and in the exit from mitosis. Interestingly, Plk4–/– cells pass through multiple divisions before they arrest at E7.5. As it is unlikely that paternal Plk4 is still present beyond E3.5 (Ref. 229), functional redundancy with other PLKs may sustain early Plk4–/– embryogenesis. Attempts to cultivate Plk4–/– embryonic fibroblasts failed, indicating that PLK4 is essential for cell viability. To explore the effects of the Plk4 gene dosage on the progression of the cell cycle in vivo, haploinsufficient Plk4+/– mice were analysed230. The elevated incidence of centrosome amplification and chromosomal instability in Plk4+/– embryonic fibroblasts correlated with the appearance of tumours in older mice.

The role of PLK1 in malignant transformation depletion of PLK1 stabilizes p53 and that the cytotoxicity The tumour suppressor p53, which is disrupted in more resulting from the depletion of PLK1 is elevated in cancer than 50% of human cancers, is of central importance in cells with defective p53 (Refs 59–61). The observation that the cellular response to DNA damage and other cellular the kinase domain of PLK1 physically binds to the DNA- stresses. Normal functioning of p53 maintains the stability binding domain of p53, thereby inhibiting both the trans- of the genome and is a potent barrier to cancer devel- activation activity and the apoptosis-inducing function of opment57,58. its activation leads to the transactivation of p53, provided the first evidence of a direct relationship target genes and the induction of growth arrest or apop- between PLK1 and p53 (Ref. 56). whether p53 is a sub- tosis, which depends on the severity of the damage in a strate of PLK1 remains controversial56,62. Nevertheless, specific cellular context. Early studies showed that the recent data describe a plausible functional link between

PLK1 and p53. it was discovered that the topoisomerase susceptibility protein (BRCA2) and CHK2 probably i-binding protein topors, which has both ubiquitin and supports mitotic progression75–77, the activity of PLK1 SuMO E3 ligase activity, is phosphorylated on Ser718 by is negatively regulated by several checkpoint pathways, PLK1 (Refs 62–64). Phosphorylation of topors by PLK1 including those involving ATM, ATR, BRCA1 and CHK1 inhibits sumoylation of p53 and simultaneously enhances (Refs 78–80). Thus, an upstream event, such as the inac- the ubiquitylation of p53. This leads to the degradation of tivation of p53, might suffice to tip the balance towards p53, and therefore supports the role of PLK1 as a nega- malignant progression by inducing a series of events that tive regulator of p53. Additional experiments showed that lead to the upregulation of the activity of PLK1, followed the long-term downregulation of PLK1 correlates with by a further weakening of the tumour suppressors. reduced levels of MDM2, a p53-specific E3 ubiquitin Although the evidence for PLK1 as an oncogene ligase that acts as a p53 antagonist65,66. This suggests that remains scarce81, accumulating evidence emphasizes the the mutual regulation of PLK1 and p53 occurs at multiple role of elevated PLK1 activity as a tumour-promoting molecular levels. it has also been shown that the transac- force. This increase in PLK1 activity stimulates mitotic tivation domain of the transcription factor p73, an addi- transcriptional programmes54; overrides the DNA- tional member of the p53 family, is phosphorylated at damage checkpoint82; contributes to checkpoint adap- Thr27 by PLK1 (Refs 67–69). Furthermore, the depletion tation83; supports the invasion through the extracellular of PLK1 in p53-deficient cells resulted in the upregula- matrix by phosphorylating vimentin84,85 and paves the tion of p73 and the pro-apoptotic protein p53-regulated way for aneuploidy as a key feature of carcinogenesis40,74. apoptosis-inducing protein 1 (p53AiP1)67. The overexpres- in parallel with these interesting insights from multiple sion of PLK1 led to reduced p73-mediated promoter experimental systems, numerous studies correlated ele- activities of the CDK inhibitor p21CiP/wAF, the cell cycle vated expression levels of PLK1 in tumour tissues with a regulator 14-3-3σ and the pro-apoptotic protein BAX68. poor prognosis, highlighting the important role of PLK1 Thus, the regulation of p73 function by PLK1 provides a in patients with cancer12,18,27,86–89. indeed, recent clinical mechanism to inhibit p53-independent apoptosis. evidence supports the theory that elevated expression A tumour-promoting role of PLK1 was also discov- of PLK1 is a general feature of human cancer90–96. PLK1 ered in an experiment designed to identify targets in inhibitors have now reached clinical development. p53-regulated pathways70. A panel of matched color- However, there is evidence that other PLK family mem- ectal cancer cell lines that differed only in the status bers have tumour suppressive roles in cancer. As there is of their endogenous p53 gene (TP53) was generated. concern about the specificity of some of the PLK1 inhibi- Remarkably, most of the genes that were upregulated tors, the biological role of other PLK family members following DNA damage in the cells without TP53 were needs to be carefully investigated. elements of the g2/M and spindle assembly checkpoints, including PLK1. interestingly, the inhibition of PLK1 The biological role of PLK2 in TP53–/– colorectal carcinoma HCT116 cells led to a As the first member of the PLK family to be identified dramatic reduction of tumours in animal models70. in mammals, PLK2 was classified as an early growth- There is also evidence for a role of PLK1 in cellular response gene by virtue of its increased expression on transformation induced by viral oncoproteins. it has been stimulation by growth factors97. whereas the expression shown that the E6 protein of the human papillomavirus profile of PLK1 supports its dominant role in mitosis, (HPv, which is implicated in the intraepithelial neopla- studies of the PLK2 transcript and protein levels in dif- sia of the lower genital tract) binds to and degrades p53 ferent cell types depict a rather complex picture. by forming a tripartite complex with the E3 ligase E6AP. First, the analysis of serum-starved, re-stimulated in addition, it has been shown that the HPv E7 protein fibroblasts and cells that re-enter g1 after mitosis showed from high-risk HPv types impairs the repressive func- that the expression of PLK2 reached a peak in early g1 tion of the retinoblastoma protein (Rb)71. PLK1 is among (Refs 97–99). Second, serum stimulation of starved pri- several g2/M regulators that are upregulated in cells that mary thyrocytes did not alter the levels of Plk2 mRNA24. express E6AP or HPv E7, and the upregulation of PLK1 Third, in HeLa cells, PLK2 was detectable during the g1, has been shown to be dependent on the degradation of S and g2 phases100. This was accompanied by a peak in p53 and the repression of pRb72. The ectopic expression the activity of PLK2 as the cells entered S phase, which of PLK1 in primary keratinocytes induced a tetraploid decreased as the cells reached g2. The intracellular state reminiscent of cells expressing E6AP or HPv E7 location of PLK2 at the centrosomes was established (Ref. 73). This indicated that the elevated levels of PLK1 in labelling experiments98,100,101. in hydroxyurea-treated induced by E6AP or HPv E7 has a principal role in the or aphidicolin-treated u2O2 osteosarcoma cells, which induction of tetraploidy, thus confirming previous studies continue to duplicate their centrioles in the absence of describing PLK1 overexpression phenotypes40,73,74. DNA synthesis, an aberrant number of centrioles were Taken together, these experiments identified a fun- induced in response to PLK2 inhibition100. Furthermore, damental role for PLK1 as a negative regulator of p53 Plk2–/– mice are viable102 (BOX 2) suggesting that PLK2 in carcinogenesis. Furthermore, studies that describe its functions are not essential or that there is some func- interplay with various other tumour suppressor proteins tional redundancy with PLK4, which has been shown to indicate that PLK1 is a central player in the development have a crucial role in the duplication of the centriole103. of human cancer. Although the PLK1-mediated con- Moreover, the ectopic expression of PLK2 induced an trol of the tumour suppressors p53, breast cancer type 2 arborized phenotype in COS-7 cells98. As the flattening

Box 3 | Cellular functions of polo-like kinase 1 (PLK1)

centrosome maturation a Centrosome maturation b G2/M transition The observation of monopolar spindles in Drosophila melanogaster cells 1 G2 phase M phase carrying a polo mutation and in cultured human cells with disrupted Microtubule PLK113 suggested that PLK1 is essential for centrosome functions and the assembly of bipolar spindles (see the figure, part a). One of these Thr14 Tyr15 CDC25 P P functions is the recruitment of the γ-tubulin ring complex (γ-TuRC), γ-TuRC Aurora A CDK1 Ser133 CDK1 which can nucleate new microtubules, to centrosomes during PLK1 P 13 prophase . Ninein-like protein (NLP), a component of the centrosome, PCM NLP PLK1 Cyclin B1 Cyclin B1 recruits γ-TuRC and stimulates microtubule nucleation during 231 interphase . PLK1 phosphorylates NLP, leading to its dissociation Kizuna WEE1/MYT1 InactiveActive from other centrosomal proteins. Centrioles Overexpression of an NLP mutant lacking PLK1 phosphorylation sites induces defects in mitotic spindle formation231. At the onset of mitosis, Cohesion of the PCM PLK1-mediated phosphorylation of NLP inhibits the dynein–dynactin- dependent targeting of NLP to the centrosome. This could represent a c Kinetochore functions and sister chromatide cohesion crucial step in the maturation process, which allows the centrosome to Prophase Metaphase SACAnaphase switch from the interphasic state to the mitotic state231,232. Centrioles PLK1 organize pericentriolar material (PCM) to form centrosomes. The EMI1 depletion of the centrosomal PLK1 substrate kizuna or the inhibition of APC/C kizuna phosphorylation by PLK1 causes fragmentation and dissociation PP2A Shugoshin 1 of the PCM from centrioles at prometaphase, resulting in multipolar Securin Kinetochore spindles and suggesting that kizuna is essential for the mitotic SCC1 233 Separase centrosome architecture . PLK1 also controls the localization of the Cohesin kinase Aurora A, which has a vital role in the function and the maturation 172,234 of centrosomes . Microtubule SA2 g2/M transition The complex of cyclin-dependent kinase 1 (CDK1) and cyclin B1, the key driver for the entry into M phase, is controlled by an intricate regulatory d Mitotic exit and cytokinesis network. On accumulation of cyclin B1 during G2, CDK1–cyclin B1 P complexes form that are kept inactive through the inhibitory MKLP2 phosphorylation of CDK1 at Thr14 and Tyr15 by the WEE1 and membrane-associated CDK1-inhibitory (MYT1) kinases (see the figure, PLK1 Self-priming GDP part b). Once the dual-specific phosphatase CDC25 dephosphorylates P and thereby activates CDK1–cyclin B1, cells enter mitosis. PLK1 seems to PRC1 ECT2 RhoA be part of a regulatory circuit that promotes the activation of CDK1 by P P Contractile ring HsCYK4 HsCYK4 assembly phosphorylating the positive regulator CDC25 and inhibiting the and contraction negative regulators WEE1 and MYT1 (Refs 235–238). In addition, cyclin MKLP1 GTP B1 is initially phosphorylated at Ser133 on centrosomes in prophase by ECT2 RhoA PLK1 (Refs 239,240). Recent data suggest that the activity of PLK1 at centrosomes is controlled by protein phosphatase 1C241. However, in inhibitor EMI1 are also located at the centrosome. In prophase, unperturbed cell divisions PLK1 might control the rate of mitotic entry, phosphorylation of EMI1 by PLK1 targets EMI1 for destruction251–253. Nature Reviews | Drug Discovery but it is not essential for this process242. The phosphorylation of APC/C subunits by PLK1 may further contribute Kinetochore functions and sister chromatide cohesion to the activation of the APC/C allowing progression beyond 254,255 In human cells various proteins (including the polo-box interacting prometaphase . One of the many APC/C substrates is securin, 256 protein (PBIB1), the inner centromere protein (INCENP), the spindle which binds to and inhibits the protease, separase . Once activated, checkpoint component BUBR1 and the mitotic checkpoint kinase separase cleaves the mitotic sister chromatid cohesion protein 1 (SCC1) BUB1) target PLK1 to the kinetochore/centromere region of subunit of the remaining centromeric cohesin protein complex, and 257,258 chromosomes37,243–245. The spindle assembly checkpoint (SAC) delays sister chromatid separation can finally take place . the transition to anaphase until all sister kinetochores experience Mitotic exit and cytokinesis bipolar attachment and are under equal tension (see the figure, part c). The spindle apparatus controls the formation of the cleavage furrow in PLK1 activity is particularly high on non-attached kinetochores the late stages of mitosis as a prerequisite for the generation of two suggesting a role of PLK1 in kinetochore attachment or in SAC physically distinct daughter cells. The spindle midzone, which is regulation246. Sister chromatids of replicated chromosomes are held situated between the two sets of separated sister chromosomes and together by a ring-shaped multi-protein complex, cohesin. In which represents a bundle of antiparallel microtubule arrays, includes a vertebrates cohesin dissociates from chromosomes in two waves. protein complex dubbed centralspindlin. Centralspindlin encompasses The vast majority is removed by the action of the so-called prophase the Rho GTPase-activating protein HsCYK4 and the kinesin motor pathway, in which PLK1 phosphorylates the SA2 subunit of cohesin247. protein MKLP1 (Refs 259,260) (see the figure, part d). PLK1 However, a small centromeric fraction of cohesin is unresponsive to this self-organizes onto the central spindle by phosphorylating and docking pathway and ensures chromosome pairing until the end of metaphase. at midzone proteins PRC1 and MKLP2 (Refs 36,38). PLK1 then A centromeric protein complex consisting of protein phosphatase 2A phosphorylates HsCYK4, thereby creating a docking site for the Rho (PP2A) and shugoshin 1 protects this subpopulation of cohesin from the GTP exchange factor ECT2 (Ref. 261). This mechanism seems to be prophase pathway and there are indications that this might occur by crucial for the assembly of an actomyosin ring at the cell cortex, which antagonizing the activity of PLK1 (Refs 248–250). The E3 ubiquitin then constricts through localized activation of the small GTPase RhoA, ligase anaphase-promoting complex/cyclosome (APC/C) and its leading to cleavage furrow formation163,262– 264.

G2 G2/M transition Mitosis P Phosphorylation of substrates Expression of bora Activation of PLK1 Degradation of bora P P P PLK1 P Bora Aurora A • Further activation of Activation of the PLK1, mitotic PLK1 by substrates CDC25, CDK1 positive bound to the PDB PLK1 kinase domain P Pre-phosphorylated feedback loop • Phosphorylation of substrate of PLK1 bound substrates PLK1 PBD domain • Targeting PLK1 to Checkpoint recovery mitotic structures Figure 2 | Model of the bora-mediated phosphorylation of PLK1 at Thr210 by Aurora A. The expression of bora peaks during G2 and decreases during mitosis48,265,266. It associates first with the inactive form of polo-like kinase 1 (PLK1). This association controls the accessibility of the activation loop of PLK1. Phosphorylation of a conserved threonine residue Nature Reviews | Drug Discovery (Thr210) in the activation loop (T-loop) of the PLK1 domain by Aurora A dramatically elevates the enzymatic activity of PLK1 at the entry into mitosis49,50. Cells in which the DNA damage checkpoint is activated arrest at the G2/M transition to allow time to repair the damage (checkpoint recovery). This process is controlled by an intricate regulatory network involving PLK1, CDC25 and CDK1 as central players (see BOX 3). The activation of the PLK1 activity through phosphorylation of Thr210 is essential for checkpoint recovery82. This mechanism seems to induce an open conformation of PLK1 and allows activated PLK1 to recruit substrates. These are primarily substrates that have been pre-phosphorylated by pro-directed kinases such as CDK1–cyclin B1. Proline phosphorylation primes these substrates for the binding of PLK1 through its polo box domain (PBD), where they undergo further phosphorylation by PLK1 during mitosis. PLK1 function is spatially regulated through the targeting activity of the conserved PBD domain.

of round mitotic cells requires massive reorganization the PLK2-controlled mitotic checkpoint represents an of the cytoskeleton, the question remains as to whether alternative mechanism by which p53 can prevent mitotic

the expression of PLK2 in early g1 and the morphologi- catastrophe and preserve genome integrity. cal effects of its overexpression indicate a role for PLK2 indeed, there are some indications that PLK2 can in the cytoskeletal reorganization process. interestingly, act as a tumour suppressor in vivo. For example, dele- the expression of PLK2 and PLK3 increased in brain tions in the long arm of chromosome 5, which are fre- neurons following stimuli that induce synaptic plastic- quent chromosomal abnormalities in acute myeloid ity104. This ability of the connection between two neu- leukaemia (AML) and human myelodysplastic syn- rons to change in strength could also be accompanied drome (a preleukaemic disorder), may contribute to by the relocation of key proteins at synapses. PLK2 and the pathogenesis of these diseases by the loss of one or PLK3 are targeted to the dendritic spines of activated two tumour suppressor genes109,110. whether this chro- neurons to mediate the phosphorylation of proteins in mosomal loss also affects the PLK2 gene, which maps this compartment104. to chromosome 5 at 5q12.1–q13.2, requires further Evidence for PLK2 as a stress-response gene came investigation. However, more direct evidence for the from a study that showed significant upregulation of tumour-suppressing function of PLK2 in haematopoi- the expression of PLK2, which was sustained for at least etic diseases was discovered in a study of B-cell malig- 24 hours in primary human thyroid cells under X-ray nancies. This study described a significantly lower irradiation24. Furthermore, microarray analysis of wild- expression of PLK2 in a wide range of B-cell neoplasms type animals and p53-null animals identified a p53- due to a Cpg methylation-dependent transcriptional dependent induction of Plk2 after irradiation in vivo105. silencing of PLK2 (Ref. 111). interestingly, the ectopic Nutlin-3a, a small-molecule inhibitor that activates the expression of PLK2 in Burkitt lymphoma cells induced p53 pathway, upregulates PLK2 in wild-type p53 cells apoptosis, indicating a selective pressure to abrogate compared with p53-null cells106. in other experiments, the function of PLK2 in this cancer111. the RNA interference (RNAi)-mediated depletion of Further support for the role of PLK2 as a tumour PLK2 in different cell lines did not alter the cell cycle suppressor was obtained from a large-scale microRNA profiles or the fraction of cells that had sub-g1 DNA (miRNA) expression profiling study in core binding content by itself. However, additional stress imposed factor AML, an AML subtype representing 10–20% of by spindle poisons led to a sensitization of the cells and primary AML112. This study identified an overexpres- was associated with a significant increase in apoptosis105. sion of miR-126/126* that was associated with a partial Thus, the depletion of PLK2 seems to phenocopy the demethylation of the Cpg island in which miR-126/126* loss of p53 by elevating the sensitivity of cells to spin- is embedded. it is thought that miRNAs contribute to dle poisons. g1 arrest can be mediated by p53 after the tumorigenesis by negatively regulating the expression cells have adapted and exited mitosis107,108. Although of their targets113. Certainly, miR-126 has been shown this particular response of the cell cycle to antimicrotu- to negatively regulate PLK2 (Ref. 112) and might thereby bule agents is dependent on the p53 target gene p21WAF1, contribute to leukaemogenesis.

The biological role of PLK3 24 hours after ionizing radiation the fibroblasts exhibited

PLK3 was originally identified and characterized as a a transcriptome with the characteristics of g0 growth fibroblast growth factor 1 (FgF1)-induced immediate quiescence, indicating that changes in gene expression, early gene with homology to other PLKs114–116. However, including the upregulation of PLK3, initiates cell cycle the pattern of PLK3 expression remains controversial. arrest. An analysis of gene expression signatures in sev- whereas some laboratories detected the PLK3 protein in eral different experiments confirmed the upregulation of every stage of the cell cycle, others observed peak levels p53 target genes, including Plk3, in response to various during mitosis or detected it in g1, but not during mito- types of stress130–133. sis116–120. A recent study suggested that these discrepancies PLK3 has been suggested as a possible tumour sup- arose primarily from the poor solubility of the endogenous pressor because it is frequently downregulated in tumours PLK3 protein and the suboptimal antibody specificity120. (lung, head and neck) and because its gene is located in a These issues also complicate the interpretation of results chromosomal region that is often involved in the loss of that describe the subcellular localization of PLK3. heterozygosity in tumour cells115,134. Mice deficient for Plk3 First, it has been identified that endogenous ectopi- are viable, but develop tumours132 (BOX 2). However, other cally expressed PLK3 localizes to the centrosomes dur- investigations of tumour tissues (breast, ovary and lung) ing interphase, to the spindle poles during mitosis, and to compared with normal tissues from the same patient did the midbody during cytokinesis121–123. Second, PLK3 has not support this model116,135,136. Comprehensive surveys also been shown to be involved in the golgi fragmenta- of large, clinically well-documented cohorts of patients tion pathway124,125. golgi membranes form stacks of cis- with cancer are required to define the role of PLK3 in solid ternae that are contained in the pericentriolar region of tumours, and the obstacles in protein detection have to be mammalian cells. During mitosis, these membranes are carefully accounted for. fragmented and dispersed throughout the cell126. it could be shown that PLK3 interacts with the protein giantin The biological role of PLK4 and the vaccinia-related kinase 1 (vRK1), suggesting From a structural point of view, PLK4 is the most diver- that PLK3 regulates golgi fragmentation during the cell gent PLK family member. There are two isoforms, SAK-a cycle124,125. Finally, PLK3 was shown to be localized only and SAK-b137. At the g1/S transition, the level of mRNA in the nucleolus and not in the centrosome, in the spindle that codes for PLK4 increases, then persists until late M poles, or in any other location in the cytoplasm120. The dis- phase and finally declines in early g1 phase138. Plk4–/– crepancies in these results may also be explained, together mice are not viable, and Plk4–/+ mice develop tumours with issues of suboptimal PLK3 antibody specificity, by (BOX 2). Experiments with a gFP–PLK4 fusion protein cell line-specific expression, post-translational modifica- showed that it localizes to the nucleolus in g2, to the cen- tions118,119 and the potential existence of different PLK3 trosomes in early M phase, to the cleavage furrow in telo- isoforms114,115. in further functional experiments, the over- phase, and to the nuclear periphery during interphase139. expression of wild-type or mutant forms of PLK3 induced As with Plk1, the overexpression of Sak-a induced multi- g2/M arrest and eventually apoptosis23,121,127. RNAi-based nucleated cells that are associated with a progressive loss depletion of PLK3, in cells that were serum-starved and of cell viability, suggesting that both genes must be regu- then re-stimulated with serum, resulted in failure of lated precisely for an ordered series of cell cycle events entry into S phase, suggesting that PLK3 is required in to occur40,138. Both gain-of-function and loss-of-function (Ref. 120) g1 . approaches clearly showed that the presence of PLK4 is interestingly, PLK3 seems to be differentially regulated crucial for the duplication of the centrioles103,140. in sum- compared with PLK1 under stressful conditions. whereas mary, although the biological roles of PLK2 and PLK3 a g2 arrest in response to DNA damage correlates with are not yet clearly defined, data describing the cellular the inhibition of PLK1 (Ref. 82), it was shown that PLK3 functions of PLK4, in particular for centriole duplication, activity rapidly increases following DNA damage in an are established. ATM-dependent manner23. This suggests that PLK3 links Chip profiling of a p53 temperature-sensitive lung can- DNA damage to cell cycle arrest through the ATM/p53 cer cell line identified PLK4 as a gene that is repressed by pathway, as documented by the enhanced association p53 (Ref. 141). Furthermore, human PLK4 was mapped of PLK3 and p53, and by the phosphorylation of p53 at to chromosome 4q27–28, a region that is frequently lost Ser20 by PLK3. Oxidative damage of DNA also triggers in hepatomas142–144. Elevated levels of PLK1 and PLK4 genotoxic stress signalling mediated by ATM22. whether were seen in a study of 71 colon tumours that were com- 145 PLK3, which is also involved in the H2O2-induced phos- pared with paired normal mucosa . As the deregulated phorylation of p53 at Ser20 (Ref. 22), lies downstream or expression of PLK4 induces multinucleation, the overex- operates in parallel to the cell cycle checkpoint kinases pression of PLK4 in tumours could contribute to chro- CHK1 and CHK2, which are downstream of the tumour mosomal instability and aneuploidy. However, the clinical suppressors ATM or ATR, to transduce stress signals is relevance of an increase in PLK4 in tumours remains to not yet known. However, in vitro experiments indicated be explored. that PLK3 is an upstream kinase of CHK2 (Ref. 119). Furthermore, when profiling the global gene expression The biological role of PLK5 of primary fibroblasts to identify early ionizing radiation- Most recently, the cloning of a fifth member of the PLK induced transcriptional responses, PLK3 was seen to be family was reported5. Murine PLK5 with a length of upregulated at 2 hours after ionizing radiation128,129. At 599 amino acids exhibits the same domain topology

as previously described for PLK1, PLK2 and PLK3. in the recent charting of the chemical space of PLK1, However, human PLK5 contains a stop codon in exon 6, various inhibitors, such as cyclapolin 1 and DAP-81, followed by a second open reading frame, which extends were identified21,161,162. They were used to explore the throughout the end of the PLK5 gene encompassing the role of PLK1 in regulating the microtubule-nucleating C-terminal third of the kinase domain and the PBD. activity of centrosomes and the balances of forces Notably, the interspecies homology between murine and that maintain spindle organization and the stability human PLK counterparts is rather low for PLK5 (57%) of kinetochore microtubules (TABLe 1). Furthermore, compared with the other family members (PLK1: 95%; ZK–thiazolidinone (TAL) represents a new chemi- PLK2: 96%; PLK3: 79%; PLK4: 78%)5. cal scaffold, which differs structurally from the PLK1 The ectopic expression of murine PLK5 induces the inhibitors previously identified161,163,164 (TABLe 1). accumulation of HEK293 and NiH3T3 cells in g1 fol- incubation with TAL induced a prometaphase-like lowed by apoptosis. The expression of PLK5, which can arrest in HeLa S3 and MCF7 cells that is characterized be seen at nucleoli, is induced by various stress-promoting by frequent monopolar spindles and MAD2-positive agents. in contrast to PLK2 and PLK3, the stress response kinetochores. This indicated that kinetochores were of PLK5 seems to be independent of p53 (Ref. 5). it will be either not attached to the spindle or not fully occu- an exciting task to explore the functions of the murine pied by microtubules and was followed by signs of cell and human PLK5 open reading frames as an example of death163. TAL treatment highlights various key func- a recent phylogenetic divergence. tions of PLK1. in TAL-treated cells the recruitment of γ-tubulin to the centrosomes is impaired; PLK1 and Issues in the design of small-molecule inhibitors Aurora A fail to localize to the centrosomes; the bipo- Cell cycle kinases like PLK1, CDKs and Aurora kinases lar spindles collapse and the metaphase plate becomes have been considered relevant drug targets owing to their disorganized in metaphase cells; sister chromatid central role in regulating the mammalian cell cycle and its arms do not separate; the functions of the cell cycle checkpoints18,21,146–149, and PLK1 has been validated as a proteins PiCH, MKLP2 and PRC1 are impaired36,38,165; drug target using various technical approaches13–16,59,150–153. and cytokinesis is impaired. Additional PLK inhibitors These insights initiated the development of rationally in preclinical development include compound 36 and designed, low molecular mass compounds that target LFM-A13, which have been tested successfully in ani- PLK1 as the Achilles heel of cancer18,21 (TABLes 1,2). mal tumour models166–169 (TABLe 1). The ATP-binding pocket of kinases is a classical tar- Rather than targeting the conserved ATP-binding get for the design of inhibitors and has been validated site, the PBD is ideally suited for studying the feasibil- by several successful strategies to inhibit members of ity of inhibiting a serine/threonine kinase by selectively the human kinome19,21,154,155. Still, the development of a influencing its protein–protein interaction. The PBD specific, ATP-competitive PLK1 inhibitor is challenging, is unique to the family of PLKs and has been shown considering the level of sequence identity of the kinase to be essential for the functions of PLK1. Recently, a domain of PLK1 with those of PLK2, PLK3 and PLK4 high-throughput screening assay based on fluorescence (53%, 54% and 37% pairwise identity, respectively)156. polarization was designed to identify small molecules The residues that line the ATP-binding pocket of human that impair the interaction of the PLK1 PBD with a PLK1 are highly conserved, with the exception of PLK4 fluorophore-labelled PBD-binding peptide170,171. This (PLK2: 90%, PLK3: 86%, PLK4: 60%)156. PLK2 and PLK3 investigation led to the identification of poloxin (a PBD display only minor differences in the residues that inter- inhibitor), the first non-peptidic inhibitor that blocks act with various inhibitors157. it is therefore conceiv- protein–protein interactions of the PLK1 PBD (TABLe 1). able that ATP-competitive inhibitors of PLK1, such as Treatment of cancer cells with poloxin induced mitotic PHA-680626 and Bi 2536, scarcely discriminate between arrest and led to a precise recapitulation of the pheno- PLK1–PLK3 (Refs 157–159). PLK4 alone was not inhib- typical effect of PBD overexpression172, whereby the ited by Bi 2536, which is probably due to the considerable localization of PLK1 was impaired (fIG. 3a,b). The core structural divergence of PLK4 compared with the other structure of poloxin is based on thymoquinone, which PLK family members. is the biologically active constituent of the volatile oil of Analysis of the crystal structure of Bi 2536 showed black seed (Nigella sativa) and is well known for its anti- that it engages in a network of interactions involving inflammatory, antioxidant and antineoplastic activi- residues from the N-terminal and C-terminal lobes of ties173,174. whether poloxin also displays antitumour the kinase domain159 (BOX 1). various regions are impor- activity is under current investigation. Further devel- tant for the potency and selectivity of Bi 2536, includ- opments led to the identification of poloxipan (a pan- ing the PLK1-specific amino acids Arg136 and Leu132 specific PBD inhibitor), which inhibits the function of (Ref. 159). One way to increase the potency and selectiv- the PBD of PLK1, PLK2 and PLK3 to similar extents ity in future structure-based drug designs would be to in vitro (Ref. 175) (TABLe 1). Remarkably, the cellular improve the interaction with the donor–acceptor pair effects of poloxipan did not differ significantly from in the hinge region while still targeting the pocket sur- those induced by poloxin. Moreover, purpurogallin, rounding Leu132 in PLK1. The crystal structure of the a natural benzotropolone compound that is extracted PLK1 domain has been determined, allowing pharma- from nutgalls obtained from trees in the genus Quercus, cophore modelling and the virtual screening of PLK1 was isolated from a screen of natural products176. inhibitors for the design of new lead structures157,159,160. Treatment of cells with purpurogallin prevented the

Table 1 | PLK 1 inhibitors in preclinical analysis

compound (chemical class) Mechanism of ic50 value for ic50/ec50 values for PLK2–PLK4 and refs action PLK1 other kinases Cyclapolin 1 (optimized benzothiazole N-oxide ) Non-competitive ~ 20 nM • Inhibits PLK1; PLK2–PLK4 were 161 O– with respect to not determined 2 R N+ O ATP; putative • Inhibits C-terminal Src kinase; covalent binding IC50 ~ 100 μM S R1 with a residue in the ATP-binding cleft NO2 DAP-81 (diaminopyrimidine (DAP) derivative) Predicted to target 0.9 μM • Inhibits PLK1; PLK2–PLK4 were not 162 N the nucleotide determined pocket • Dose-dependent reduction of HN NH2 N CDC25C phosphorylation in cells and NH recapitulation of key aspects of the loss-of-function phenotype for PLK1 O HN O Ph

ZK-thiazolidinone (ZK-thiazolidinone) ATP-competitive 19 ± 12 nM • High selectivity for PLK1 in a panel 163 F F inhibitor of 93 serine/threonine and tyrosine kinases O F NH • PLK2 and PLK3 were inhibited at S IC <100 nM N N 50 H • No significant inhibition of PLK4 N O N Compound 36 (imidazopyridine derivative) Not determined 9.8 nM • High selectivity for PLK1 in a panel of 166 212 kinases CONH2 • PLK2 IC50 = 21 nM; PLK3 IC50 = 178 nM; K2 S EC = 19 nM O 50

N NH N O F F

LFM-A13 Molecular docking Plx1 32.5 μM • PLK3 IC50 = 61 μM 167– studies predict using GST-CDC25 • Also inhibits human BTK with an IC 169 OH O 50 X binding to the as a substrate of 17.2 ± 0.8 μM N catalytic site of the H Plx1 kinase domain CN

Poloxin (thymoquinone derivative) Interferes with PLK1 PLK1 PBD: 4.8 ± • PBDs of PLK2 IC50 = 18.7 ± 1.8 μM; 170 O PBD functions in 1.3 μM PLK3 IC50 = 53.9 ± 8.5 μM vitro and in vivo • Poloxin does not significantly affect other O subtypes of the phosphothreonine/phos- N phoserine-binding domains (FHA domain of CHK2, WW domain of PIN1) or the phosphotyrosine-binding domains (SH2 O domains of STAT1, STAT3, STAT5 and LCK) Poloxipan Pan-specific • PLK1 PBD: • Poloxipan inhibits other subtypes of 175 inhibitor of the PBDs 3.2 ± 0.3 μM the phosphothreonine/phosphoserine- CH3 of PLK1, PLK2 and binding domains (FHA domain of O O PLK3 CHK2, WW domain of PIN1) and the phosphotyrosine-binding domains (SH2 N N S domains of STAT1, STAT3, STAT5 and LCK) to a significantly lesser extent compared N Br to PBDs of PLKs (PLK2 PDB IC = 1.7 ± O 50 0.2 μM; PLK3 PDB IC50 =3.0 ± 0.1 μM) Purpurogallin (benzotropolone-containing Inhibits ~ 0.3 µM in • Also inhibits the PBD of PLK2, but not 176 compound) PBD-dependent GST-pulldown of PLK3 OH O binding in vitro and assays using PLK1 OH in vivo PBD as bait for HO WEE1A

BTK, Bruton’s tyrosine kinase; EC50, effector concentration for half-maximum response; FHA, forkhead-associated domain; GST, glutathione-S-transferase; IC50, half-maximal inhibitory concentration; LCK, lymphocyte-specific protein tyrosine kinase; PBD, polo box domain; PIN1, peptidylprolyl cis/trans isomerase, NIMA-interacting 1; PLK, polo-like kinase; Plx1, Xenopus homologue of PLK1; SH2, Src homology 2; STAT, signal transducer and activator of transcription.

Table 2 | PLK1 inhibitors in clinical development

compound (chemical class) Mechanism of ic50 value for PLK1 ic50/ec50 values for PLK2– refs action PLK4 and other kinases ON 01910.Na (benzylstyryl sulphone) Affects microtubule 9–10 nM; no inhibition • Multitargeted inhibitor 157, 181, H CO dynamics of PLK1 up to 30 μM 183–188 3 in vitro

HO OCH3 O2S HOOCH2CHN H3CO

HMN-176 (stilbenazole compound) Alters spatial • Not determined • Not determined 189–192 distribution of PLK1 (no direct catalytic N O inhibition) H3C NH O SO2

GSK461364 (thiophene derivative) ATP-competitive EC50 < 50 nM in > 83% • Has 400-fold greater 193–201 N O inhibitor of the 120 cancer cell potency for PLK1 than for S N lines tested PLK2 and PLK3 NH2 Ki = 2.2 nM

N N F C H3C 3

NMS-P937 (pyrazoloquinazoline scaffold) ATP-competitive < 0.02 μM • Shows high selectivity in a 203 inhibitor panel of more than 250 protein N kinases • PLK2 IC = 10 μM R3 N R2 50 • PLK3 IC50 = 3.2 μM N N R1 BI 2536 (dihydropteridinone derivative) ATP-competitive 0.83 nM • Shows at least 1,000-fold 158, 182,

H3C inhibitor selectivity against a panel of 204–209 O tyrosine and serine/threonine H N N kinases

H • PLK2 IC50 = 3.5 nM N N CH3 • PLK3 IC50 = 9.0 nM N • EC = 2–25 nM N O N 50 H3C O

BI 6727 (dihydropteridinone derivative) ATP-competitive 0.87 nM • No inhibitory activity against a 210, 211 O inhibitor panel of more than 50 protein HN kinases • PLK2 IC50 = 5 nM • PLK3 IC50 = 56 nM • EC = 11–37 nM N CH3 50 N O HN N

CH3 N O N N

EC50, effector concentration for half-maximum response; IC50, half-maximal inhibitory concentration; Ki, inhibition dissociation constant; PLK, polo-like kinase.

localization of PLK1 to the centrosomes and induced a if future experiments support the essential role of chromosome congression phenotype in mitotic cells176. PLK2 and PLK3 as tumour suppressors in cancer, PBD As purpurogallin impairs proteins that are unrelated inhibitors may be more suited for cancer therapy, since to PBD, such as tyrosine kinases, Hiv-1 integrase and they are more selective for PLK1 versus PLK2 and PLK3 prolyl endopeptidases (half-maximal inhibitory con- than the ATP-competitive compounds.

centration (iC50) values of 27.5 μM, 2 μM and 16 μM, As the list of PLK inhibitors is growing rapidly, owing respectively), the chemical nature of these interactions to tremendous pharmaceutical efforts, only a selection needs to be analysed in order to define the inhibitory of new compounds tested in preclinical settings can be mechanism of action177–179. presented here (TABLe 1).

a DAPI α-tubulin γ-tubulin Merged beneficial to target non-structural, crucial components of mitotic cells, including kinesins and protein kinases like PLK1. Several inhibitors targeting PLKs are cur- DMSO rently in development and are under investigation in a growing number of clinical trials.

ON 01910.Na. ON 01910.Na inhibits the growth of cancer cells and induces apoptosis in multiple cancer cell lines in vitro and in vivo181 (TABLe 2). ON 01910. Poloxin Na-treated cells exhibit multipolar spindles, accumulate at the g2/M phase and exhibit spindle and centrosome abnormalities. whether this result represents a loss-of- function phenotype for PLK1, which is dominant over the inhibition of other non-related kinases that are targeted by ON 01910.Na, is still controversial181,182. An additional TQ investigation showed that ON 01910.Na did not inhibit PLK1 up to a concentration of 30 μM in in vitro kinase assays and it did not cause cellular phenotypes that were reminiscent of PLK1 RNAi157. A recent, first-in-human Phase i study in adult b DAPI α-tubulin γ-tubulin Merged patients with solid tumours defined the maximum tolerated dose (MTD) as 3,120 mg of ON 01910.Na when administered as a 2-hour infusion, twice weekly for 3 weeks, in 4-week cycles183. The most common adverse DMSO events included fatigue, pain, nausea, vomiting and abdominal pain. Examination of the pharmacokinetics showed that the drug was distributed rapidly and that the elimination phase was relatively slow. Additional Phase i studies for the treatment of patients suffering Poloxin from advanced cancers or B-cell chronic lymphocytic leukaemia are currently underway184–188.

HMN-176. HMN-176, which is an active metabolite of the synthetic antitumour compound HMN-214, interferes with the subcellular localization of PLK1, induces TQ g2/M arrest and downregulates the expression of MDR1 by inhibiting the activity of NF-Y189,190 (TABLe 2). HMN- 214 shows antitumour activity against various human tumour xenografts. Two dose escalation trials of HMN- Figure 3 | chromosome congression defects and the mislocalization of PLK1 by polo box domain inhibition. a | HeLa cells in mitosis were stained with 4′,6-diamidino- 214 in patients with advanced solid tumours were con- Nature Reviews | Drug Discovery 2-phenylindole (DAPI) or labelled using α-tubulin-specific or γ-tubulin-specific ducted to determine the MTD and the pharmacokinetics antibodies for analysis by microscopy. A merged image of the labelled cells was created. of the compound190–192. Both studies showed antitumour HeLa cells treated with dimethyl sulphoxide (DMSO) were used as a control. In the activity, accompanied by modest adverse effects (myal- presence of several doses of poloxin and thymoquinone (TQ) a dose-dependent increase gia/bone pain syndrome, hyperglycaemia, neutropenic of HeLa cells arrested in prometaphase was observed. Among the cells that were sepsis and neuropathy) that depended on the schedule arrested in a metaphase-like state, a dose-dependent increase of cells in which single or of oral dosing and the degree of pretreatment. multiple chromosomes had failed to congress to the metaphase plate was noticed170. b | While in metaphase polo-like kinase 1 (PLK1) located to centrosomes and kinetochores in DMSO-treated control cells. Both centrosomal and kinetochoral localization of PLK1 GW843682X and GSK461364. gw843682X, also called were significantly reduced in the presence of TQ and poloxin. Reproduced, with compound 1, and gSK461364 resulted from chemical permission, from Ref. 170. Bar represents 25 μm. optimization studies of thiophene as a lead structure, which had been shown to have a good PLK1 inhibitory activity164,193,194. These studies included docking into a Clinical development homology model of PLK1 and structure–activity relation- Drugs like the vinca alkaloids, taxanes and the recently ship exploration of different thiophene derivatives64,193–195 identified first member of the epothilone family that (TABLe 2). Although gw843682X induced g2/M arrest hamper the dynamic activity of microtubules, have and apoptosis in different types of cancer cells, which proved to be successful in the clinical treatment of can- express abundant PLK1, gSK461364 also showed favour- cer180. However, some of their adverse effects, such as able anticancer activity in vivo196–200. in an ongoing first- neurotoxicity and myelosuppression, may be attrib- in-human study of gSK461364 in patients with advanced uted to interactions with the tubulin cytoskeleton in solid tumours, the primary objective is to determine the non-dividing differentiated cells. Hence, it might be MTD and the pharmacokinetics of the compound201. As

a secondary objective, the study will examine the anti- excretion and toxicity showed good oral availability. The tumour activity of gSK461364 (Ref. 201). Patients who first-in-human Phase i dose-escalation study included 50 received escalating doses of gSK461364 in different sched- patients with advanced or metastatic solid tumours, each ules showed different dose-limiting toxicity effects, includ- of whom received Bi 6727 at doses of 12 mg to 450 mg211. ing sepsis at a dose of 225 mg or pulmonary embolism and Encouraging antitumour activity was accompanied by neutropaenia at a dose of 100 mg201. limited, reversible and treatable adverse haematological events, including neutropaenia and thrombocytopaenia. NMS-P937. Structure-driven drug design on a pyrazoloquinazoline scaffold led to the discovery of NMS-P937 Future perspectives (Ref. 202) (TABLe 2). NMS-P937 induced g2/M arrest and Cancer cells undergo various types of cellular stress (DNA showed antiproliferative activity against more than 100 cell damage, proteotoxic, mitotic, metabolic and oxidative lines from haematological and solid cancers. This com- stress)26 and require stress support pathways for their pound, which is available in oral form, is well tolerated and survival. Beyond the activation of a crucial transcrip- efficacious in preclinical models203. it has recently entered tional programme required for mitotic progression54, clinical trials. PLK1 seems to improve the tolerance of cells to cancer- associated cellular stress. For example, it was shown that BI 2536. The ATP-competitive PLK1 inhibitor Bi 2536 was the overexpression of a hyperactive PLK1 mutant in cells identified by screening a diverse library of organic com- harbouring damaged DNA overrides the DNA-damage pounds. it causes prometaphase arrest, mitotic abnormali- checkpoint82. PLK1 also improves the tolerance towards ties and apoptosis in a wide range of tumour cell lines158,182 the mitotic stress associated with chromosome instability (TABLe 2). in a high percentage of treated cells, monopolar (CiN phenotype)92. spindles are formed. These are reminiscent of the spin- A recent study used 74,905 retroviral short hairpin dle phenotype observed in polo mutants of Drosophila1. RNAs (shRNAs), targeting 32,293 transcripts, to iden- Remarkably, Bi 2536 showed high efficacy in vivo at well- tify the vulnerabilities of cancers that have oncogenic tolerated doses and caused tumour regression in several mutations in the small gTPase Ras212. This investigation xenograft models158. showed that cells in which mutant Ras acts as an oncogene Three Phase i dose-escalation studies with different are highly sensitive to PLK1 inhibition, because mitotic schedules were conducted to determine the MTD, safety stress seems to be a key event in these cells. The inhibition and tolerability, pharmacokinetics and antitumour effect of PLK1 seems to induce stress sensitization by downregu- of Bi 2536 in patients with advanced and/or metastatic lating the activity of the stress support pathways, leading tumours204–206. Bi 2536 was seen to be well tolerated. The ultimately to apoptosis of cancer cells. whereas cancer cells MTD was 200 mg, when used once a week at 200 mg per may not be able to cope with cancer-associated stresses fol- cycle or when used twice a week at 100 mg per cycle. lowing PLK1 silencing, non-transformed cells seem to be The main dose-limiting toxicity was reversible neutro- less dependent on PLK1 expression15,60. However, it will paenia with or without infection204. Nausea, fatigue and be important to explore the dependence of primary non- anorexia were of moderate to mild intensity. Although transformed cells on PLK1. a partial response was a rare event, 23% of the patients increasing evidence documents that PLK2 and PLK3 experienced stable disease for 3 months or more, at become activated under stresses such as ultraviolet irra- doses equal to or above the MTD. Patients suffering diation or hypoxia. The stimulation of PLK2 and PLK3 as from non-small cell lung cancer, pancreatic cancer or executioners of p53 safety functions may slow or even stop hormone-refractory prostate cancer were included in mitotic progression, giving normal cells a chance to com- three Phase ii trials with Bi 2536 (Refs 207–209). Bi 2536 pensate for stressful situations or to prevent further stress was generally well tolerated and neutropaenia was the overload. At least in haematological cancers, the inactiva- main side effect. Fatigue and nausea were also observed, tion of PLK2 implies a tumour suppressor function for generally as mild adverse events. The clinical benefit PLK2. Obviously, cancer cells have developed strategies to provides a strong rationale for further clinical develop- abandon important safety features, including the functions ment of this drug. of PLK2 and PLK3 downstream of p53, thereby relieving the checkpoints that usually prevent aberrant growth and BI 6727. Screening of a diverse library of small molecules genomic instability. Previous studies on cancer cells and for their ability to inhibit the catalytic activity of PLK1 led tissues lack a comprehensive analysis of the activities of to the identification of a series of dihydropteridinones210. all PLK family members. Thus, considering the tremen- Subsequent lead optimization facilitated the synthesis of dous importance of p53 for human cancer, it will be of the a range of dihydropteridinone analogues with improved utmost importance to focus not only on the activities of parameters, which included Bi 6727 (Ref. 210) (TABLe 2). PLK1, but to simultaneously consider the activities of the Analysis of the co-crystal with the T210v mutant PLK1 other PLK family members. This will improve our under- kinase domain placed Bi 6727 into the ATP-binding standing of the mutual regulatory network of p53 and the pocket210. Bi 6727, which induced prometaphase arrest PLKs in cancer. The interactomes of PLK2, PLK3 and accompanied by monopolar spindles and the phospho- PLK4 should be studied further to complete the picture of rylation of histone H3 at Ser10, showed high efficacy the signal transduction pathways initiated by PLKs. against various cancer cell lines in culture and in vivo. Because antimitotic drugs belong to the most suc- Analysis of absorption, distribution, metabolism, cessful classes of anticancer drugs, the development of

small-molecule inhibitors targeting cardinal mitotic PLK inhibitors. However, we still do not know whether regulators like PLK1 is a promising strategy to fight can- the tumours observed in these mouse models reflect cer. increasing clinical evidence supports the evaluation the crucial roles of PLK1, PLK3 and PLK4 in adult cells of certain PLK1-specific compounds at the Phase ii/iii or point to their essential roles during embryogenesis, level. Thus, it will be essential to monitor the early tumour when the elevated activity of PLKs is required for rapid response to these drugs in primary tissues. The recently proliferation and/or differentiation. To shed new light on developed, sensitive enzyme-linked immunosorbent this important issue, new mouse models that allow for assay (ELiSA)-based PLK1 assay, which can quantify the the inhibition of PLKs exclusively during adulthood are activity of PLK1 in small volumes of cellular lysates, will essential. However, although PLK4 does not seem to be become a valuable clinical tool that can determine the an off target for PLK1 inhibitors because of its structural effect of new inhibitors in primary tissues213. Combining divergence, various new compounds are promiscuous this assay with additional biomarkers for the assessment in vitro as they also partially inhibit PLK2 and PLK3. of mitotic abnormalities and apoptosis should be priori- This unwanted side effect in stressed cancer cells could tized, so that optimal treatment schedules using PLK1 support malignant progression in long-term clinical inhibitors can be developed. applications and when using an uncontrolled dosage. As A fascinating issue will be to determine whether ATP- long as we do not know whether the inhibition of PLK1 competitive and PBD-specific inhibitors of PLK1 differ in is always dominant over the inhibition of PLK2 and their efficacy to suppress tumour growth. understanding PLK3, the development and use of biomarkers for the the precise nature of the PBD-binding of new compounds entire PLK family is a priority for the development of new like poloxin, poloxipan and purpurogallin will be impor- PLK1 inhibitors. The inclusion of pharmacodynamic end tant for optimizing specificity and providing new insights points that show target inhibition into clinical trials will into the design of additional PBD inhibitors. Poloxin may support the definition of optimal administration schemes bind covalently to surface residues in the PLK1 PBD. The and schedules that can operate in parallel to the currently concern that this putative mode of action of poloxin causes established pharmacokinetic monitoring. unacceptable levels of cytotoxicity in vivo was not substan- Still, multiple questions remain. Do these new tiated in our xenograft experiments (J. Yuan et al., personal compounds induce the enzymatic inhibition of PLK1 communication). Before advancing to the clinical stage, a accompanied by mitotic catastrophe and apoptosis in critical assessment of the efficacy and the toxicity of new the tumour cells of treated patients? How do the effects PBD inhibitors in different animal models of cancer is induced by PLK1 inhibitors compare to those of tra- required. Alternative strategies for the inhibition of PLK1, ditional antitubulin cytotoxics? is the p53 status a key including the use of chemically modified small interfer- determinant for the therapeutic response of tumour ing RNA or specific albumin nanoparticles, are currently cells to PLK1 inhibitors in cancer patients? An improved being developed214–217. understanding of the concerted action of different PLKs Although small molecules that target PLK1 hold great in haematopoietic and solid cancers, and an under- promise as anticancer drugs, the increased tumour devel- standing of how the inhibition of PLK1 leads to cell opment that is observed in Plk1+/–, Plk3–/– and Plk4+/– mice death, are essential to the future success of PLK1-based may raise concerns with regards to the safety of novel interventions.

1. Sunkel, C. E. & Glover, D. M. polo, a mitotic mutant 10. Archambault, V. & Glover, D. M. Polo-like kinases: 16. Spankuch-Schmitt, B., Bereiter-Hahn, J., Kaufmann, M. of Drosophila displaying abnormal spindle poles. conservation and divergence in their functions and & Strebhardt, K. Effect of RNA silencing of polo-like J. Cell Sci. 89 (Pt 1), 25–38 (1988). regulation. Nature Rev. Mol. Cell Biol. 10, 265–275 kinase 1 (PLK1) on apoptosis and spindle formation This paper describes the mitotic phenotype of cells (2009). in human cancer cells. J. Natl. Cancer Inst. 94, homozygous for mutant alleles of the locus polo. 11. Barr, F. A. & Gruneberg, U. Cytokinesis: placing 1863–1877 (2002). 2. Llamazares, S. et al. polo encodes a protein kinase and making the final cut. Cell 131, 847–860 17. Liu, X. & Erikson, R. L. Activation of Cdc2/cyclin B and homolog required for mitosis in Drosophila. Genes (2007). inhibition of centrosome amplification in cells depleted Dev. 5, 2153–2165 (1991). References 3, 4, 7–11 are excellent reviews on of Plk1 by siRNA. Proc. Natl Acad. Sci. USA 99, 3. Glover, D. M., Hagan, I. M. & Tavares, A. A. Polo-like the multiple roles of PLK1 and its phylogenetic 8672–8676 (2002). kinases: a team that plays throughout mitosis. Genes counterparts during mitosis. 18. Strebhardt, K. & Ullrich, A. Targeting polo-like Dev. 12, 3777–3787 (1998). 12. Holtrich, U. et al. Induction and down-regulation of kinase 1 for cancer therapy. Nature Rev. Cancer 6, 4. Barr, F. A., Sillje, H. H. & Nigg, E. A. Polo-like kinases PLK, a human serine/threonine kinase expressed in 321–330 (2006). and the orchestration of cell division. Nature Rev. Mol. proliferating cells and tumors. Proc. Natl Acad. Sci. 19. Strebhardt, K. & Ullrich, A. Paul Ehrlich’s magic bullet Cell Biol. 5, 429–440 (2004). USA 91, 1736–1740 (1994). concept: 100 years of progress. Nature Rev. Cancer 8, 5. Andrysik, Z. et al. The novel mouse polo-like kinase 5 This report describes for the first time the elevated 473–480 (2008). responds to DNA damage and localizes in the levels of PLK1 in human cancer and initiated many 20. Schoffski, P. Polo-like kinase (PLK) inhibitors in nucleolus. Nucleic Acids Res. 38, 2931–2943 (2010). follow-on studies analysing the expression preclinical and early clinical development in oncology. 6. Clay, F. J., McEwen, S. J., Bertoncello, I., Wilks, A. F. signature of PLK1 in a broad spectrum of human Oncologist 14, 559–570 (2009). & Dunn, A. R. Identification and cloning of a protein tumours. 21. McInnes, C., Mezna, M. & Fischer, P. M. Progress in kinase-encoding mouse gene, Plk, related to the polo 13. Lane, H. A. & Nigg, E. A. Antibody microinjection the discovery of polo-like kinase inhibitors. Curr. Top. gene of Drosophila. Proc. Natl Acad. Sci. USA 90, reveals an essential role for human polo-like Med. Chem. 5, 181–197 (2005). 4882–4886 (1993). kinase 1 (Plk1) in the functional maturation of 22. Xie, S. et al. Reactive oxygen species-induced 7. van de Weerdt, B. C. & Medema, R. H. Polo-like mitotic centrosomes. J. Cell Biol. 135, 1701–1713 phosphorylation of p53 on serine 20 is mediated in kinases: a team in control of the division. Cell Cycle 5, (1996). part by polo-like kinase 3. J. Biol. Chem. 276, 853–864 (2006). 14. Cogswell, J. P., Brown, C. E., Bisi, J. E. & Neill, S. D. 36194–36199 (2001). 8. Petronczki, M., Lenart, P. & Peters, J. M. Polo on the Dominant-negative polo-like kinase 1 induces mitotic 23. Xie, S. et al. Plk3 functionally links DNA damage to cell rise — from mitotic entry to cytokinesis with Plk1. catastrophe independent of cdc25C function. Cell cycle arrest and apoptosis at least in part via the p53 Dev. Cell 14, 646–659 (2008). Growth Differ. 11, 615–623 (2000). pathway. J. Biol. Chem. 276, 43305–43312 (2001). 9. Takaki, T., Trenz, K., Costanzo, V. & Petronczki, M. 15. Spankuch-Schmitt, B. et al. Downregulation of human 24. Shimizu-Yoshida, Y. et al. Radiation-inducible hSNK Polo-like kinase 1 reaches beyond mitosis — polo-like kinase activity by antisense oligonucleotides gene is transcriptionally regulated by p53 binding cytokinesis, DNA damage response, and development. induces growth inhibition in cancer cells. Oncogene homology element in human thyroid cells. Biochem. Curr. Opin. Cell Biol. 20, 650–660 (2008). 21, 3162–3171 (2002). Biophys. Res. Commun. 289, 491–498 (2001).

25. Hanahan, D. & Weinberg, R. A. The hallmarks of 47. Ellinger-Ziegelbauer, H. et al. Ste20-like kinase (SLK), 72. Patel, D., Incassati, A., Wang, N. & McCance, D. J. cancer. Cell 100, 57–70 (2000). a regulatory kinase for polo-like kinase (Plk) during the Human papillomavirus type 16 E6 and E7 cause 26. Luo, J., Solimini, N. L. & Elledge, S. J. Principles G2/M transition in somatic cells. Genes Cells 5, polyploidy in human keratinocytes and up-regulation of cancer therapy: oncogene and non-oncogene 491–498 (2000). of G2-M-phase proteins. Cancer Res. 64, 1299–1306 addiction. Cell 136, 823–837 (2009). 48. Chan, E. H., Santamaria, A., Sillje, H. H. & Nigg, E. A. (2004). 27. Yuan, J. et al. Polo-like kinase, a novel marker for Plk1 regulates mitotic Aurora A function through 73. Incassati, A., Patel, D. & McCance, D. J. Induction of cellular proliferation. Am. J. Pathol. 150, 1165–1172 βTrCP-dependent degradation of hBora. Chromosoma tetraploidy through loss of p53 and upregulation of (1997). 117, 457–469 (2008). Plk1 by human papillomavirus type-16 E6. Oncogene 28. Lowery, D. M. et al. Proteomic screen defines the 49. Seki, A., Coppinger, J. A., Jang, C. Y., Yates, J. R. & 25, 2444–2451 (2006). Polo-box domain interactome and identifies Rock2 Fang, G. Bora and the kinase Aurora A cooperatively 74. Meraldi, P., Honda, R. & Nigg, E. A. Aurora-A as a Plk1 substrate. EMBO J. 26, 2262–2273 activate the kinase Plk1 and control mitotic entry. overexpression reveals tetraploidization as a major (2007). Science 320, 1655–1658 (2008). route to centrosome amplification in p53-/- cells. 29. Lee, K. S., Grenfell, T. Z., Yarm, F. R. & Erikson, R. L. 50. Macurek, L. et al. Polo-like kinase-1 is activated by EMBO J. 21, 483–492 (2002). Mutation of the polo-box disrupts localization and Aurora A to promote checkpoint recovery. Nature 75. Lin, H. R., Ting, N. S., Qin, J. & Lee, W. H. M phase- mitotic functions of the mammalian polo kinase Plk. 455, 119–123 (2008). specific phosphorylation of BRCA2 by polo-like kinase 1 Proc. Natl Acad. Sci. USA 95, 9301–9306 (1998). References 48–50 establish the role of bora for correlates with the dissociation of the BRCA2-P/CAF This report describes the first example of a the activation of PLK1 by Aurora A. complex. J. Biol. Chem. 278, 35979–35987 (2003). functional analysis of the polo-box showing that 51. Laoukili, J. et al. FoxM1 is required for execution of 76. Lee, M., Daniels, M. J. & Venkitaraman, A. R. this unique module regulates the subcellular the mitotic programme and chromosome stability. Phosphorylation of BRCA2 by the polo-like kinase localization of PLK1. Nat. Cell Biol. 7, 126–136 (2005). Plk1 is regulated by DNA damage and mitotic 30. Lee, K. S., Song, S. & Erikson, R. L. The polo-box- 52. Major, M. L., Lepe, R. & Costa, R. H. Forkhead box progression. Oncogene 23, 865–872 (2004). dependent induction of ectopic septal structures by M1B transcriptional activity requires binding of Cdk- 77. Tsvetkov, L., Xu, X., Li, J. & Stern, D. F. Polo-like kinase a mammalian polo kinase, plk, in Saccharomyces cyclin complexes for phosphorylation-dependent 1 and Chk2 interact and co-localize to centrosomes cerevisiae. Proc. Natl Acad. Sci. USA 96, recruitment of p300/CBP coactivators. Mol. Cell Biol. and the midbody. J. Biol. Chem. 278, 8468–8475 14360–14365 (1999). 24, 2649–2661 (2004). (2003). 31. Song, S., Grenfell, T. Z., Garfield, S., Erikson, R. L. & 53. Wang, I. C. et al. Forkhead box M1 regulates the 78. van Vugt, M. A., Smits, V. A., Klompmaker, R. & Lee, K. S. Essential function of the polo box of Cdc5 in transcriptional network of genes essential for mitotic Medema, R. H. Inhibition of polo-like kinase-1 by DNA subcellular localization and induction of cytokinetic progression and genes encoding the SCF (Skp2-Cks1) damage occurs in an ATM- or ATR-dependent fashion. structures. Mol. Cell Biol. 20, 286–298 (2000). ubiquitin ligase. Mol. Cell Biol. 25, 10875–10894 J. Biol. Chem. 276, 41656–41660 (2001). 32. Seong, Y. S. et al. A spindle checkpoint arrest and a (2005). 79. Ree, A. H., Bratland, A., Nome, R. V., Stokke, T. & cytokinesis failure by the dominant-negative polo-box 54. Fu, Z. et al. Plk1-dependent phosphorylation of Fodstad, O. Repression of mRNA for the PLK cell cycle domain of Plk1 in U-2 OS cells. J. Biol. Chem. 277, FoxM1 regulates a transcriptional programme gene after DNA damage requires BRCA1. Oncogene 32282–32293 (2002). required for mitotic progression. Nat. Cell Biol. 10, 22, 8952–8955 (2003). 33. Reynolds, N. & Ohkura, H. Polo boxes form a single 1076–1082 (2008). 80. Tang, J., Erikson, R. L. & Liu, X. Checkpoint kinase 1 functional domain that mediates interactions with 55. Martin, B. T. & Strebhardt, K. Polo-like kinase 1: (Chk1) is required for mitotic progression through multiple proteins in fission yeast polo kinase. J. Cell target and regulator of transcriptional control. negative regulation of polo-like kinase 1 (Plk1). Sci. 116, 1377–1387 (2003). Cell Cycle 5, 2881–2885 (2006). Proc. Natl Acad. Sci USA 103, 11964–11969 (2006). 34. Elia, A. E., Cantley, L. C. & Yaffe, M. B. Proteomic 56. Ando, K. et al. Polo-like kinase 1 (Plk1) inhibits p53 81. Smith, M. R. et al. Malignant transformation of screen finds pSer/pThr-binding domain localizing Plk1 function by physical interaction and phosphorylation. mammalian cells initiated by constitutive expression of to mitotic substrates. Science 299, 1228–1231 J. Biol. Chem. 279, 25549–25561 (2004). the polo-like kinase. Biochem. Biophys. Res. Commun. (2003). 57. Whibley, C., Pharoah, P. D. & Hollstein, M. p53 234, 397–405 (1997). This excellent paper identifies the polo-box polymorphisms: cancer implications. Nature Rev. 82. Smits, V. A. et al. Polo-like kinase-1 is a target of domain of Plk1 as a specific phosphoserine Cancer 9, 95–107 (2009). the DNA damage checkpoint. Nature Cell Biol. 2, or phosphothreonine binding domain. 58. Vazquez, A., Bond, E. E., Levine, A. J. & Bond, G. L. 672–676 (2000). 35. van de Weerdt, B. C. et al. Polo-box domains confer The genetics of the p53 pathway, apoptosis and This paper reports that PLK1 is an important target specificity to the polo-like kinase family. cancer therapy. Nature Rev. Drug Discov. 7, 979–987 target of the DNA damage checkpoint. Biochim. Biophys. Acta 1783, 1015–1022 (2008). (2008). 83. Syljuasen, R. G., Jensen, S., Bartek, J. & Lukas, J. 36. Neef, R. et al. Phosphorylation of mitotic kinesin-like 59. Liu, X. & Erikson, R. L. Polo-like kinase (Plk)1 Adaptation to the ionizing radiation-induced G2 protein 2 by polo-like kinase 1 is required for depletion induces apoptosis in cancer cells. Proc. Natl checkpoint occurs in human cells and depends on cytokinesis. J. Cell Biol. 162, 863–875 (2003). Acad. Sci. USA 100, 5789–5794 (2003). checkpoint kinase 1 and polo-like kinase 1 kinases. 37. Kang, Y. H. et al. Self-regulated Plk1 recruitment to 60. Liu, X., Lei, M. & Erikson, R. L. Normal cells, but not Cancer Res. 66, 10253–10257 (2006). kinetochores by the Plk1-PBIP1 interaction is critical cancer cells, survive severe Plk1 depletion. Mol. Cell 84. Yamaguchi, T. et al. Phosphorylation by Cdk1 induces for proper chromosome segregation. Mol. Cell 24, Biol. 26, 2093–2108 (2006). Plk1-mediated vimentin phosphorylation during 409–422 (2006). 61. Guan, R. et al. Small interfering RNA-mediated mitosis. J. Cell Biol. 171, 431–436 (2005). 38. Neef, R. et al. Choice of Plk1 docking partners polo-like kinase 1 depletion preferentially reduces the 85. Rizki, A., Mott, J. D. & Bissell, M. J. Polo-like kinase 1 during mitosis and cytokinesis is controlled by the survival of p53-defective, oncogenic transformed cells is involved in invasion through extracellular matrix. activation state of Cdk1. Nat. Cell Biol. 9, 436–444 and inhibits tumor growth in animals. Cancer Res. 65, Cancer Res. 67, 11106–11110 (2007). (2007). 2698–2704 (2005). 86. Eckerdt, F., Yuan, J. & Strebhardt, K. Polo-like kinases 39. Jang, Y. J., Lin, C. Y., Ma, S. & Erikson, R. L. 62. Yang, X. et al. Plk1-mediated phosphorylation of and oncogenesis. Oncogene 24, 267–276 (2005). Functional studies on the role of the C-terminal topors regulates p53 stability. J. Biol. Chem. 284, 87. Takai, N., Hamanaka, R., Yoshimatsu, J. & Miyakawa, I. domain of mammalian polo-like kinase. Proc. Natl 18588–18592 (2009). Polo-like kinases (Plks) and cancer. Oncogene 24, Acad. Sci. USA 99, 1984–1989 (2002). 63. Weger, S., Hammer, E. & Heilbronn, R. Topors acts as 287–291 (2005). 40. Mundt, K. E., Golsteyn, R. M., Lane, H. A. & a SUMO-1 E3 ligase for p53 in vitro and in vivo. FEBS 88. Wolf, G. et al. Prognostic significance of polo-like Nigg, E. A. On the regulation and function of human Lett. 579, 5007–5012 (2005). kinase (PLK) expression in non-small cell lung cancer. polo-like kinase 1 (PLK1): effects of overexpression 64. Rajendra, R. et al. Topors functions as an E3 ubiquitin Oncogene 14, 543–549 (1997). on cell cycle progression. Biochem. Biophys. Res. ligase with specific E2 enzymes and ubiquitinates p53. 89. Knecht, R. et al. Prognostic significance of polo-like Commun. 239, 377–385 (1997). J. Biol. Chem. 279, 36440–36444 (2004). kinase (PLK) expression in squamous cell carcinomas 41. Elia, A. E. et al. The molecular basis for 65. Momand, J., Wu, H. H. & Dasgupta, G. MDM2-master of the head and neck. Cancer Res. 59, 2794–2797 phosphodependent substrate targeting and regulator of the p53 tumor suppressor protein. Gene (1999). regulation of Plks by the polo-box domain. Cell 242, 15–29 (2000). References 88 and 89 both describe for the first 115, 83–95 (2003). 66. Kreis, N. N. et al. Long-term downregulation of polo- time the clinical relevance of PLK1 overexpression 42. Golsteyn, R. M., Mundt, K. E., Fry, A. M. & Nigg, E. A. like kinase 1 increases the cyclin-dependent kinase for patients with tumours. Cell cycle regulation of the activity and subcellular inhibitor p21WAF1/CIP1. Cell Cycle 8, 460–472 (2009). 90. Salvatore, G. et al. A cell proliferation and localization of Plk1, a human protein kinase 67. Koida, N. et al. Inhibitory role of Plk1 in the regulation chromosomal instability signature in anaplastic implicated in mitotic spindle function. J. Cell Biol. of p73-dependent apoptosis through physical thyroid carcinoma. Cancer Res. 67, 10148–10158 129, 1617–1628 (1995). interaction and phosphorylation. J. Biol. Chem. 283, (2007). 43. Jang, Y. J., Ma, S., Terada, Y. & Erikson, R. L. 8555–8563 (2008). 91. Liu, L., Zhang, M. & Zou, P. Expression of PLK1 and Phosphorylation of threonine 210 and the role of 68. Soond, S. M. et al. p73-mediated transcriptional survivin in diffuse large B-cell lymphoma. Leuk. serine 137 in the regulation of mammalian polo-like activity is negatively regulated by polo-like kinase 1. Lymphoma 48, 2179–2183 (2007). kinase. J. Biol. Chem. 277, 44115–44120 (2002). Cell Cycle 7, 1214–1223 (2008). 92. Yamamoto, Y. et al. Overexpression of polo-like 44. Kelm, O., Wind, M., Lehmann, W. D. & Nigg, E. A. 69. Kaghad, M. et al. Monoallelically expressed gene kinase 1 (PLK1) and chromosomal instability in Cell cycle-regulated phosphorylation of the related to p53 at 1p36, a region frequently deleted in bladder cancer. Oncology 70, 231–237 (2006). Xenopus polo-like kinase Plx1. J. Biol. Chem. 277, neuroblastoma and other human cancers. Cell 90, 93. Liu, L., Zhang, M. & Zou, P. Polo-like kinase 1 as a 25247–25256 (2002). 809–819 (1997). new target for non-Hodgkin’s lymphoma treatment. 45. Daub, H. et al. Kinase-selective enrichment enables 70. Sur, S. et al. A panel of isogenic human cancer cells Oncology 74, 96–103 (2008). quantitative phosphoproteomics of the kinome across suggests a therapeutic approach for cancers with 94. Kanaji, S. et al. Expression of polo-like kinase 1 (PLK1) the cell cycle. Mol. Cell 31, 438–448 (2008). inactivated p53. Proc. Natl Acad. Sci. USA 106, protein predicts the survival of patients with gastric 46. Qian, Y. W., Erikson, E. & Maller, J. L. Purification and 3964–3969 (2009). carcinoma. Oncology 70, 126–133 (2006). cloning of a protein kinase that phosphorylates and 71. zur Hausen, H. Papillomaviruses and cancer: from 95. Jang, Y. J., Kim, Y. S. & Kim, W. H. Oncogenic effect activates the polo-like kinase Plx1. Science 282, basic studies to clinical application. Nature Rev. of polo-like kinase 1 expression in human gastric 1701–1704 (1998). Cancer 2, 342–350 (2002). carcinomas. Int. J. Oncol. 29, 589–594 (2006).

96. Weichert, W. et al. Expression patterns of polo-like 121. Wang, Q. et al. Cell cycle arrest and apoptosis induced 145. Macmillan, J. C., Hudson, J. W., Bull., S., Dennis, J. W. kinase 1 in human gastric cancer. Cancer Sci. 97, by human polo-like kinase 3 is mediated through & Swallow, C. J. Comparative expression of the 271–276 (2006). perturbation of microtubule integrity. Mol. Cell Biol. mitotic regulators SAK and PLK in colorectal cancer. 97. Simmons, D. L., Neel, B. G., Stevens, R., Evett, G. 22, 3450–3459 (2002). Ann. Surg. Oncol. 8, 729–740 (2001). & Erikson, R. L. Identification of an early-growth- 122. Jiang, N., Wang, X., Jhanwar-Uniyal, M., 146. Keen, N. & Taylor, S. Aurora-kinase inhibitors as response gene encoding a novel putative protein Darzynkiewicz, Z. & Dai, W. Polo box domain of anticancer agents. Nature Rev. Cancer 4, 927–936 kinase. Mol. Cell Biol. 12, 4164–4169 (1992). Plk3 functions as a centrosome localization signal, (2004). 98. Ma, S., Liu, M. A., Yuan, Y. L. & Erikson, R. L. The overexpression of which causes mitotic arrest, 147. Pérez de Castro, I., de Cárcer, G., Montoya, G. & serum-inducible protein kinase Snk is a G1 phase cytokinesis defects, and apoptosis. J. Biol. Chem. Malumbres, M. Emerging cancer therapeutic polo-like kinase that is inhibited by the calcium- and 281, 10577–10582 (2006). opportunities by inhibiting mitotic kinases. Curr. Opin. integrin-binding protein CIB. Mol. Cancer Res. 1, 123. Yang, F. et al. Identification of a novel mitotic Pharmacol. 8, 375–383 (2008). 376–384 (2003). phosphorylation motif associated with protein 148. McInnes, C. Progress in the evaluation of CDK 99. Anger, M. et al. Cell cycle dependent expression localization to the mitotic apparatus. J. Cell Sci. inhibitors as anti-tumor agents. Drug Discov. Today of Plk1 in synchronized porcine fetal fibroblasts. 120, 4060–4070 (2007). 13, 875–881 (2008). Mol. Reprod. Dev. 65, 245–253 (2003). 124. Ruan, Q. et al. Polo-like kinase 3 is golgi localized and 149. Lapenna, S. & Giordano, A. Cell cycle kinases as 100. Warnke, S. et al. Polo-like kinase-2 is required for involved in regulating golgi fragmentation during the therapeutic targets for cancer. Nature Rev. Drug centriole duplication in mammalian cells. Curr. Biol. cell cycle. Exp. Cell Res. 294, 51–59 (2004). Discov. 8, 547–566 (2009). 14, 1200–1207 (2004). 125. Lopez-Sanchez, I., Sanz-Garcia, M. & Lazo, P. A. Plk3 150. Yuan, J., Kramer, A., Eckerdt, F., Kaufmann, M. & 101. Cizmecioglu, O., Warnke, S., Arnold, M., Duensing, S. interacts with and specifically phosphorylates VRK1 Strebhardt, K. Efficient internalization of the polo-box & Hoffmann, I. Plk2 regulated centriole duplication is in Ser342, a downstream target in a pathway that of polo-like kinase 1 fused to an antennapedia peptide dependent on its localization to the centrioles and a induces golgi fragmentation. Mol. Cell Biol. 29, results in inhibition of cancer cell proliferation. Cancer functional polo-box domain. Cell Cycle 7, 3548–3555 1189–1201 (2009). Res. 62, 4186–4190 (2002). (2008). 126. Sutterlin, C., Hsu, P., Mallabiabarrena, A. & Malhotra, V. 151. Spankuch, B. et al. Cancer inhibition in nude mice 102. Ma, S., Charron, J. & Erikson, R. L. Role of Plk2 (Snk) Fragmentation and dispersal of the pericentriolar after systemic application of U6 promoter-driven short in mouse development and cell proliferation. Mol. Cell golgi complex is required for entry into mitosis in hairpin RNAs against PLK1. J. Natl. Cancer Inst. 96, Biol. 23, 6936–6943 (2003). mammalian cells. Cell 109, 359–369 (2002). 862–872 (2004). 103. Habedanck, R., Stierhof, Y. D., Wilkinson, C. J. & 127. Conn, C. W., Hennigan, R. F., Dai, W., Sanchez, Y. & 152. Matthess, Y. et al. Conditional inhibition of cancer cell Nigg, E. A. The polo kinase Plk4 functions in centriole Stambrook, P. J. Incomplete cytokinesis and induction proliferation by tetracycline-responsive, H1 promoter- duplication. Nature Cell Biol. 7, 1140–1146 (2005). of apoptosis by overexpression of the mammalian driven silencing of PLK1. Oncogene 24, 2973–2980 104. Kauselmann, G. et al. The polo-like protein kinases polo-like kinase, Plk3. Cancer Res. 60, 6826–6831 (2005). Fnk and Snk associate with a Ca2+- and integrin- (2000). 153. Kappel, S., Matthess, Y., Zimmer, B., Kaufmann, M. binding protein and are regulated dynamically with 128. Zhou, T. et al. Profiles of global gene expression in & Strebhardt, K. Tumor inhibition by genomically synaptic plasticity. EMBO J. 18, 5528–5539 (1999). ionizing-radiation-damaged human diploid fibroblasts integrated inducible RNAi-cassettes. Nucleic Acids 105. Burns, T. F., Fei, P., Scata, K. A., Dicker, D. T. & reveal synchronization behind the G1 checkpoint in a Res. 34, 4527–4536 (2006). El-Deiry, W. S. Silencing of the novel p53 target gene G0-like state of quiescence. Environ. Health Perspect. 154. Manning, G., Whyte, D. B., Martinez, R., Hunter, T. Snk/Plk2 leads to mitotic catastrophe in paclitaxel 114, 553–559 (2006). & Sudarsanam, S. The protein kinase complement (taxol)-exposed cells. Mol. Cell Biol. 23, 5556–5571 129. Kis, E. et al. Microarray analysis of radiation response of the human genome. Science 298, 1912–1934 (2003). genes in primary human fibroblasts. Int. J. Radiat. (2002). 106. Tovar, C. et al. Small-molecule MDM2 antagonists Oncol. Biol. Phys. 66, 1506–1514 (2006). 155. Goldstein, D. M., Gray, N. S. & Zarrinkar, P. P. reveal aberrant p53 signaling in cancer: implications 130. Staib, F. et al. The p53 tumor suppressor network High-throughput kinase profiling as a platform for for therapy. Proc. Natl Acad. Sci. USA 103, is a key responder to microenvironmental components drug discovery. Nature Rev. Drug Discov. 7, 391–397 1888–1893 (2006). of chronic inflammatory stress. Cancer Res. 65, (2008). 107. Khan, S. H. & Wahl, G. M. p53 and pRb prevent 10255–10264 (2005). 156. Johnson, E. F., Stewart, K. D., Woods, K. W., rereplication in response to microtubule inhibitors by 131. Han, E. S. et al. The in vivo gene expression signature Giranda, V. L. & Luo, Y. Pharmacological and mediating a reversible G1 arrest. Cancer Res. 58, of oxidative stress. Physiol. Genomics 34, 112–126 functional comparison of the polo-like kinase family: 396–401 (1998). (2008). insight into inhibitor and substrate specificity. 108. Lanni, J. S. & Jacks, T. Characterization of the p53- 132. Yang, Y. et al. Polo-like kinase 3 functions as a tumor Biochemistry 46, 9551–9563 (2007). dependent postmitotic checkpoint following spindle suppressor and is a negative regulator of hypoxia- 157. Kothe, M. et al. Structure of the catalytic domain disruption. Mol. Cell Biol. 18, 1055–1064 (1998). inducible factor-1α under hypoxic conditions. Cancer of human polo-like kinase 1. Biochemistry 46, 109. Van den Berghe, H. & Michaux, L. 5q, twenty-five Res. 68, 4077–4085 (2008). 5960–5971 (2007). years later: a synopsis. Cancer Genet. Cytogenet. 133. Wang, L., Gao, J., Dai, W. & Lu, L. Activation of 158. Steegmaier, M. et al. BI 2536, a potent and selective 94, 1–7 (1997). polo-like kinase 3 by hypoxic stresses. J. Biol. Chem. inhibitor of polo-like kinase 1, inhibits tumor growth 110. Nimer, S. D. & Golde, D. W. The 5q– abnormality. 283, 25928–25935 (2008). in vivo. Curr. Biol. 17, 316–322 (2007). Blood 70, 1705–1712 (1987). 134. Dai, W. et al. PRK, a cell cycle gene localized to 8p21, This report laid the foundation for the preclinical 111. Syed, N. et al. Transcriptional silencing of polo-like is downregulated in head and neck cancer. Genes and clinical testing of the promising compound kinase 2 (SNK/PLK2) is a frequent event in B-cell Chromosomes. Cancer 27, 332–336 (2000). BI 2536. malignancies. Blood 107, 250–256 (2006). 135. Weichert, W. et al. Polo-like kinase isoforms in breast 159. Kothe, M. et al. Selectivity-determining residues in 112. Li, Z. et al. Distinct microRNA expression profiles in cancer: expression patterns and prognostic Plk1. Chem. Biol. Drug Des. 70, 540–546 (2007). acute myeloid leukemia with common translocations. implications. Virchows Arch. 446, 442–450 (2005). 160. Wang, H. Y. et al. Pharmacophore modeling and virtual Proc. Natl Acad. Sci. USA 105, 15535–15540 136. Weichert, W. et al. Polo-like kinase isoform screening for designing potential PLK1 inhibitors. Bioorg. (2008). expression is a prognostic factor in ovarian Med. Chem. Lett. 18, 4972–4977 (2008). 113. Esquela-Kerscher, A. & Slack, F. J. Oncomirs – carcinoma. Br. J. Cancer 90, 815–821 (2004). 161. McInnes, C. et al. Inhibitors of polo-like kinase reveal microRNAs with a role in cancer. Nature Rev. Cancer 137. Fode, C., Motro, B., Yousefi, S., Heffernan, M. & roles in spindle-pole maintenance. Nature Chem. Biol. 6, 259–269 (2006). Dennis, J. W. Sak, a murine protein-serine/threonine 2, 608–617 (2006). 114. Donohue, P. J., Alberts, G. F., Guo, Y. & Winkles, J. A. kinase that is related to the Drosophila polo kinase 162. Peters, U., Cherian, J., Kim, J. H., Kwok, B. H. & Identification by targeted differential display of an and involved in cell proliferation. Proc. Natl Acad. Kapoor, T. M. Probing cell-division phenotype space immediate early gene encoding a putative serine/ Sci. USA 91, 6388–6392 (1994). and polo-like kinase function using small molecules. threonine kinase. J. Biol. Chem. 270, 10351–10357 138. Fode, C., Binkert, C. & Dennis, J. W. Constitutive Nature Chem. Biol. 2, 618–626 (2006). (1995). expression of murine Sak-a suppresses cell growth References 161 and 162 are both excellent 115. Li, B. et al. Prk, a cytokine-inducible human protein and induces multinucleation. Mol. Cell Biol. 16, papers that describe the effects of novel serine/threonine kinase whose expression appears to 4665–4672 (1996). small-molecule inhibitors targeting the kinase be down-regulated in lung carcinomas. J. Biol. Chem. 139. Hudson, J. W. et al. Late mitotic failure in mice lacking domain of PLK1. 271, 19402–19408 (1996). Sak, a polo-like kinase. Curr. Biol. 11, 441–446 163. Santamaria, A. et al. Use of the novel Plk1 inhibitor 116. Holtrich, U. et al. Adhesion induced expression of the (2001). ZK-thiazolidinone to elucidate functions of Plk1 in serine/threonine kinase Fnk in human macrophages. 140. Bettencourt-Dias, M. et al. SAK/PLK4 is required early and late stages of mitosis. Mol. Biol. Cell 18, Oncogene 19, 4832–4839 (2000). for centriole duplication and flagella development. 4024–4036 (2007). 117. Ouyang, B. et al. Human Prk is a conserved protein Curr. Biol. 15, 2199–2207 (2005). 164. Lansing, T. J. et al. In vitro biological activity of a serine/threonine kinase involved in regulating 141. Li, J. et al. SAK, a new polo-like kinase, is novel small-molecule inhibitor of polo-like kinase 1. M phase functions. J. Biol. Chem. 272, transcriptionally repressed by p53 and induces Mol. Cancer Ther. 6, 450–459 (2007). 28646–28651 (1997). apoptosis upon RNAi silencing. Neoplasia. 7, 165. Baumann, C., Korner, R., Hofmann, K. & Nigg, E. A. 118. Chase, D. et al. Expression and phosphorylation of 312–323 (2005). PICH, a centromere-associated SNF2 family ATPase, fibroblast-growth-factor-inducible kinase (Fnk) during 142. Deloukas, P. et al. A physical map of 30,000 human is regulated by Plk1 and required for the spindle cell-cycle progression. Biochem. J. 333 (Pt 3), genes. Science 282, 744–746 (1998). checkpoint. Cell 128, 101–114 (2007). 655–660 (1998). 143. Hammond, C., Jeffers, L., Carr, B. I. & Simon, D. 166. Sato, Y. et al. Imidazopyridine derivatives as potent 119. Bahassi el, M. et al. Mammalian polo-like kinase 3 Multiple genetic alterations, 4q28, a new suppressor and selective polo-like kinase (PLK) inhibitors. Bioorg. (Plk3) is a multifunctional protein involved in stress region, and potential gender differences in human Med. Chem. Lett. 19, 4673–4678 (2009). response pathways. Oncogene 21, 6633–6640 hepatocellular carcinoma. Hepatology 29, 167. Mahajan, S. et al. Rational design and synthesis of a (2002). 1479–1485 (1999). novel anti-leukemic agent targeting Bruton’s tyrosine 120. Zimmerman, W. C. & Erikson, R. L. Polo-like kinase 3 144. Hudson, J. W., Chen, L., Fode, C., Binkert, C. & kinase (BTK), LFM-A13 [α-cyano-β-hydroxy-β-methyl-N- is required for entry into S phase. Proc. Natl Acad. Dennis, J. W. Sak kinase gene structure and (2,5-dibromophenyl)propenamide]. J. Biol. Chem. Sci. USA 104, 1847–1852 (2007). transcriptional regulation. Gene 241, 65–73 (2000). 274, 9587–9599 (1999).

168. Uckun, F. M. et al. Anti-breast cancer activity of 189. Tanaka, H. et al. HMN-176, an active metabolite of 206. Ellis, P. M. et al. A phase I dose escalation trial of BI LFM-A13, a potent inhibitor of polo-like kinase (PLK). the synthetic antitumor agent HMN-214, restores 2536, a novel Plk1 inhibitor, with standard dose Bioorg. Med. Chem. 15, 800–814 (2007). chemosensitivity to multidrug-resistant cells by pemetrexed in previously treated advanced or 169. Uckun, F. M. Chemosensitizing anti-cancer activity targeting the transcription factor NF-Y. Cancer Res. metastatic non-small cell lung cancer (NSCLC). of LFM-A13, a leflunomide metabolite analog 63, 6942–6947 (2003). J. Clin. Oncol. 26, 8115 (2008). targeting polo-like kinases. Cell Cycle 6, 3021–3026 190. Garland, L. L., Taylor, C., Pilkington, D. L., Cohen, J. L. 207. von Pawel, J. et al. Randomized phase II trial of two (2007). & Von Hoff, D. D. A phase I pharmacokinetic study of dosing schedules of BI 2536, a novel Plk-1 inhibitor, 170. Reindl, W., Yuan, J., Kramer, A., Strebhardt, K. & HMN-214, a novel oral stilbene derivative with polo- in patients with relapsed advanced or metastatic non- Berg, T. Inhibition of polo-like kinase 1 by blocking like kinase-1-interacting properties, in patients with small-cell lung cancer (NSCLC). J. Clin. Oncol. 26, polo-box domain-dependent protein–protein advanced solid tumors. Clin. Cancer Res. 12, 8030 (2008). interactions. Chem. Biol. 15, 459–466 (2008). 5182–5189 (2006). 208. Pandha, H. S. et al. An open label phase II trial of This paper describes the inhibition of the 191. Von Hoff, D. D., Taylor, C., Rubin, S., Cohen, J. & BI 2536, a novel Plk1 inhibitor, in patients with polo-box domain function by a small-molecule Garland, L. A phase I and pharmacokinetic study of metastatic hormone refractory prostate cancer inhibitor. HMN-214, a novel oral polo-like kinase inhibitor, in (HRPC). J. Clin. Oncol. 26, 14547 (2008). 171. Reindl, W., Strebhardt, K. & Berg, T. A high- patients with advanced solid tumors. J. Clin. Oncol. 209. Mross, K. et al. A randomized phase II trial of the throughput assay based on fluorescence 22 (Suppl. 14), 3034 (2004). novel polo-like kinase 1 inhibitor BI 2536 in polarization for inhibitors of the polo-box domain of 192. Patnaik, A. et al. HMN-214, a novel oral chemonaïve patients with unresectable advanced polo-like kinase 1. Anal. Biochem. 383, 205–209 antimicrotubular agent and inhibitor of polo-like- and pancreatic cancer: a study in cooperation with the (2008). cyclin-dependent kinases: clinical, pharmacokinetic CESAR network of investigators. Ann. Oncol. 19 172. Hanisch, A., Wehner, A., Nigg, E. A. & Sillje, H. H. (PK) and pharmacodynamic (PD) relationships (Suppl. 8), 493P (2008). Different Plk1 functions show distinct dependencies on observed in a phase I trial of a daily x 5 schedule 210. Rudolph, D. et al. BI 6727, a polo-like kinase inhibitor polo-box domain-mediated targeting. Mol. Biol. Cell every 28 days. Proc. Am. Soc. Clin. Oncol. 22, 514 with improved pharmacokinetic profile and broad 17, 448–459 (2006). (2003). antitumor activity. Clin. Cancer Res. 15, 3094–3102 173. Gali-Muhtasib, H., Roessner, A. & Schneider-Stock, R. 193. Emmitte, K. A. et al. Discovery of thiophene inhibitors (2009). Thymoquinone: a promising anti-cancer drug from of polo-like kinase. Bioorg. Med. Chem. Lett. 19, 211. Schöffski, P. et al. A phase I single dose escalation natural sources. Int. J. Biochem. Cell Biol. 38, 1018–1021 (2009). study of the novel polo-like kinase I inhibitor BI 6727 1249–1253 (2006). 194. Emmitte, K. A. et al. Design of potent thiophene in patients with advanced solid tumors. Eur. J. Cancer 174. Kaseb, A. O. et al. Androgen receptor and E2F-1 inhibitors of polo-like kinase 1 with improved solubility 6 (Suppl.), 14–15 (2008). targeted thymoquinone therapy for hormone- and reduced protein binding. Bioorg. Med. Chem. 212. Luo, J. et al. A genome-wide RNAi screen identifies refractory prostate cancer. Cancer Res. 67, Lett. 19, 1694–1697 (2009). multiple synthetic lethal interactions with the Ras 7782–7788 (2007). 195. Erskine, S. et al. Biochemical characterization of oncogene. Cell 137, 835–848 (2009). 175. Reindl, W., Yuan, J., Kramer, A., Strebhardt, K. & GSK461364: a novel, potent, and selective inhibitor 213. Park, J. E. et al. Direct quantification of polo-like Berg, T. A pan-specific inhibitor of the polo-box of polo-like kinase 1 (Plk1). Proc. Annu. Meet. Am. kinase 1 activity in cells and tissues using a highly domains of polo-like kinases arrests cancer cells in Assoc. Cancer Res. Abstr. 3257 (2007). sensitive and specific ELISA assay. Proc. Natl Acad. mitosis. Chembiochem. 10, 1145–1148 (2009). 196. Laquerre, S. et al. A potent and selective polo-like Sci. USA 106, 1725–1730 (2009). 176. Watanabe, N. et al. Deficiency in chromosome kinase 1 (Plk1) inhibitor (GSK461364) induces cell 214. Judge, A. D. et al. Confirming the RNAi-mediated congression by the inhibition of Plk1 polo box cycle arrest and growth inhibition of cancer cell. mechanism of action of siRNA-based cancer domain-dependent recognition. J. Biol. Chem. 284, AACR meeting website [online], http://www. therapeutics in mice. J. Clin. Invest. 119, 661–673 2344–2353 (2009). aacrmeetingabstracts.org/cgi/gca?allch=&SEARCHID (2009). 177. Abou-Karam, M. & Shier, W. T. Inhibition of oncogene =1&AUTHOR1=Laquerre&FIRSTINDEX=0&hits=1 215. Spankuch, B. et al. Downregulation of Plk1 expression product enzyme activity as an approach to cancer 0&RESULTFORMAT=&gca=aacrmtg%3B2007%2 by receptor-mediated uptake of antisense chemoprevention. Tyrosine-specific protein kinase F1_Annual_Meeting%2F5389 (2007). oligonucleotide-loaded nanoparticles. Neoplasia. 10, inhibition by purpurogallin from Quercus sp. nutgall. 197. Erskine, S. et al. Biochemical characterization of 223–234 (2008). Phytother. Res. 13, 337–340 (1999). GSK461364: a novel, potent, and selective inhibitor 216. Steinhauser, I. M., Langer, K., Strebhardt, K. M. & 178. Farnet, C. M. et al. Human immunodeficiency virus of polo-like kinase-1 (Plk1). AACR meeting website Spankuch, B. Effect of trastuzumab-modified antisense type 1 cDNA integration: new aromatic hydroxylated [online], http://www.aacrmeetingabstracts.org/cgi/gca? oligonucleotide-loaded human serum albumin inhibitors and studies of the inhibition mechanism. allch=&SEARCHID=1&AUTHOR1=Erskine&FULLTE nanoparticles prepared by heat denaturation. Antimicrob. Agents Chemother. 42, 2245–2253 XT=Plk1&FIRSTINDEX=0&hits=10&RESULTFORM Biomaterials 29, 4022–4028 (2008). (1998). AT=&gca=aacrmtg%3B2007%2F1_Annual_ 217. Steinhauser, I., Langer, K., Strebhardt, K. & 179. Inamori, Y. et al. Biological activity of purpurogallin. Meeting%2F3257 (2007). Spankuch, B. Uptake of plasmid-loaded nanoparticles Biosci. Biotechnol. Biochem. 61, 890–892 (1997). 198. Sutton, D. et al. Efficacy of GSK461364, a selective in breast cancer cells and effect on Plk1 expression. 180. Jackson, J. R., Patrick, D. R., Dar, M. M. & Huang, Plk1 inhibitor, in human tumor xenograft models. J. Drug Target. 17, 627–637 (2009). P. S. Targeted anti-mitotic therapies: can we improve AACR meeting website [online], http://www. 218. Bandeiras, T. M. et al. Structure of wild-type Plk-1 on tubulin agents? Nature Rev. Cancer 7, 107–117 aacrmeetingabstracts.org/cgi/gca?allch=&SEARCHID kinase domain in complex with a selective DARPin. (2007). =1&AUTHOR1=Sutton&FULLTEXT=GSK461364& Acta Crystallogr. D. Biol. Crystallogr. 64, 339–353 181. Gumireddy, K. et al. ON01910, a non-ATP-competitive FIRSTINDEX=0&hits=10&RESULTFORMAT=&gca (2008). small molecule inhibitor of Plk1, is a potent anticancer =aacrmtg%3B2007%2F1_Annual_ 219. Nolen, B., Taylor, S. & Ghosh, G. Regulation of protein agent. Cancer Cell 7, 275–286 (2005). Meeting%2F5388 (2007). kinases; controlling activity through activation segment 182. Lenart, P. et al. The small-molecule inhibitor BI 2536 199. Ikezoe, T. et al. A novel treatment strategy targeting conformation. Mol. Cell 15, 661–675 (2004). reveals novel insights into mitotic roles of polo-like polo-like kinase 1 in hematological malignancies. 220. Cheng, K. Y., Lowe, E. D., Sinclair, J., Nigg, E. A. & kinase 1. Curr. Biol. 17, 304–315 (2007). Leukemia 23, 1564–1576 (2009). Johnson, L. N. The crystal structure of the human 183. Jimeno, A. et al. Phase I study of ON 01910.Na, a 200. Didier, C., Cavelier, C., Quaranta, M., Demur, C. & polo-like kinase-1 polo box domain and its novel modulator of the polo-like kinase 1 pathway, Ducommun, B. Evaluation of polo-like kinase 1 phospho-peptide complex. EMBO J. 22, 5757–5768 in adult patients with solid tumors. J. Clin. Oncol. inhibition on the G2/M checkpoint in acute myelocytic (2003). 26, 5504–5510 (2008). leukaemia. Eur. J. Pharmacol. 591, 102–105 (2008). 221. García-Alvarez, B., de Cárcer, G., Ibañez, S., Bragado- 184. Jimeno, A. et al. Evaluation of the novel mitotic 201. Olmos, D. et al. Phase I first-in-human study of the Nilsson, E. & Montoya, G. Molecular and structural modulator ON 01910.Na in pancreatic cancer and polo-like kinase-1 selective inhibitor, GSK461364, basis of polo-like kinase 1 substrate recognition: preclinical development of an ex vivo predictive assay. in patients with advanced solid tumors. J. Clin. Oncol. implications in centrosomal localization. Proc. Natl Oncogene 28, 610–618 (2009). 27 (Suppl. 15) 3536 (2009). Acad. Sci. USA 104, 3107–3112 (2007). 185. Weinstein, J. et al. Phase I study of ON 01910.Na, 202. Beria, I. et al. Identification of 4, 5-Dihydro-1H- 222. Yun, S. M. et al. Structural and functional analyses of a novel polo-like kinase 1 pathway modulator, pyrazolo[4,3-h]quinazoline derivatives as a new class minimal phosphopeptides targeting the polo-box administered as a weekly 24-hour continuous infusion of orally and selective polo-like kinase 1 inhibitors. domain of polo-like kinase 1. Nature Struct. Mol. Biol. in patients with advanced cancer. J. Clin. Oncol. 26, J. Med. Chem. 53 (Suppl. 9), 3532–3551 16, 876–882 (2009). (2008). (2010). 223. Leung, G. C. et al. The Sak polo-box comprises a 186. Ghalib,M. H. et al. ON01910.Na, a novel polo-like 203. Valsasina, B. et al. Pyrazoloquinazolines: from an structural domain sufficient for mitotic subcellular kinase pathway modulator as a treatment for patients unselective hit to a potent Plk1-specific inhibitor. localization. Nature Struct. Biol. 9, 719–724 (2002). with advanced cancer. AACR meeting website [online], AACR website [online], http://www.abstractsonline. 224. Lu, L. Y. et al. Polo-like kinase 1 is essential for early http://www.aacrmeetingabstracts.org/cgi/gca?allch= com/viewer/viewAbstractPrintFriendly. embryonic development and tumor suppression. &SEARCHID=1&AUTHOR1=Ghalib&FIRSTINDEX= asp?CKey={716D6BF1-9F4C-475D-BC07- Mol. Cell Biol. 28, 6870–6876 (2008). 0&hits=10&RESULTFORMAT=&gca=aacrmtg%3B 49C954251C47}&SKey={8479D63F-A150-4BA5- This excellent report describes the structure and 2008%2F1_Annual_Meeting%2F5652 (2008). A046-74836CFE9ECB}&MKey={D007B270-E8F6- function of the PLK4 polo-box. 187. Ohnuma, T. et al. Phase I study of ON 01910.Na by 492D-803B-7582CE7A0988}&AKey={728BCE9C- 225. Kitada, K., Johnson, A. L., Johnston, L. H. & Sugino, A. 3-day continuous infusion (CI) in patients (pts) with 121B-46B9-A8EE-DC51FDFC6C15} (2009). A multicopy suppressor gene of the Saccharomyces advanced cancer. J. Clin. Oncol. 24 (Suppl. 18), 204. Mross, K. et al. Phase I dose escalation and cerevisiae G1 cell cycle mutant gene dbf4 encodes a 13137 (2006). pharmacokinetic study of BI 2536, a novel polo-like protein kinase and is identified as CDC5. Mol. Cell 188. Chapman, C. M., Perez-Galan, P. & Wiestner, A. kinase 1 inhibitor, in patients with advanced solid Biol. 13, 4445–4457 (1993). ON 01910.Na, a novel clinical grade PLK-1 inhibitor, tumors. J. Clin. Oncol. 26, 5511–5517 (2008). 226. Ohkura, H., Hagan, I. M. & Glover, D. M. The selectively induces apoptosis in human B-cell chronic 205. Hofheinz, R. et al. A phase I repeated dose escalation conserved Schizosaccharomyces pombe kinase plo1, lymphocytic leukemia (B-CLL). Onconova website study of the polo-like kinase 1 inhibitor BI 2536 in required to form a bipolar spindle, the actin ring, and [online], http://www.onconova.com/pdfs/publications/ patients with advanced solid tumors. J. Clin. Oncol. septum, can drive septum formation in G1 and G2 AACR2009%20Abstracts.pdf (2009). 24 (Suppl 18), 2038 (2006). cells. Genes Dev. 9, 1059–1073 (1995).

227. Liby, K. et al. Identification of the human homologue 243. Elowe, S., Hummer, S., Uldschmid, A., Li, X. & 258. Hornig, N. C. & Uhlmann, F. Preferential cleavage of the early-growth response gene Snk, encoding a Nigg, E. A. Tension-sensitive Plk1 phosphorylation of chromatin-bound cohesin after targeted serum-inducible kinase. DNA Seq. 11, 527–533 on BubR1 regulates the stability of kinetochore phosphorylation by polo-like kinase. EMBO J. 23, (2001). microtubule interactions. Genes Dev. 21, 2205–2219 3144–3153 (2004). 228. Karn, T. et al. Human SAK related to the PLK/polo (2007). 259. Mishima, M., Kaitna, S. & Glotzer, M. Central spindle family of cell cycle kinases shows high mRNA 244. Goto, H. et al. Complex formation of Plk1 and INCENP assembly and cytokinesis require a kinesin-like expression in testis. Onc. Rep. 4, 505–510 (1997). required for metaphase–anaphase transition. Nat. Cell protein/RhoGAP complex with microtubule bundling 229. Nothias, J. Y., Majumder, S., Kaneko, K. J. & Biol. 8, 180–187 (2006). activity. Dev. Cell 2, 41–54 (2002). DePamphilis, M. L. Regulation of gene expression at 245. Qi, W., Tang, Z. & Yu, H. Phosphorylation- and polo- 260. Mishima, M., Pavicic, V., Gruneberg, U., Nigg, E. A. the beginning of mammalian development. J. Biol. box-dependent binding of Plk1 to Bub1 is required for & Glotzer, M. Cell cycle regulation of central spindle Chem. 270, 22077–22080 (1995). the kinetochore localization of Plk1. Mol. Biol. Cell 17, assembly. Nature 430, 908–913 (2004). 230. Ko, M. A. et al. Plk4 haploinsufficiency causes mitotic 3705–3716 (2006). 261. Burkard, M. E. et al. Plk1 self-organization and infidelity and carcinogenesis. Nature Genet. 37, 246. Ahonen, L. J. et al. Polo-like kinase 1 creates the priming phosphorylation of HsCYK-4 at the spindle 883–888 (2005). tension-sensing 3F3/2 phosphoepitope and modulates midzone regulate the onset of division in human cells. 231. Casenghi, M. et al. Polo-like kinase 1 regulates the association of spindle-checkpoint proteins at PLoS. Biol. 7, e1000111 (2009). Nlp, a centrosome protein involved in microtubule kinetochores. Curr. Biol. 15, 1078–1089 (2005). 262. Brennan, I. M., Peters, U., Kapoor, T. M. & Straight, A. F. nucleation. Dev. Cell 5, 113–125 (2003). 247. Sumara, I. et al. The dissociation of cohesin from Polo-like kinase controls vertebrate spindle elongation 232. Casenghi, M., Barr, F. A. & Nigg, E. A. Phosphorylation chromosomes in prophase is regulated by polo-like and cytokinesis. PLoS. One. 2, e409 (2007). of Nlp by Plk1 negatively regulates its dynein- kinase. Mol. Cell 9, 515–525 (2002). 263. Burkard, M. E. et al. Chemical genetics reveals the dynactin-dependent targeting to the centrosome. 248. Riedel, C. G. et al. Protein phosphatase 2A protects requirement for polo-like kinase 1 activity in positioning J. Cell Sci. 118, 5101–5108 (2005). centromeric sister chromatid cohesion during RhoA and triggering cytokinesis in human cells. 233. Oshimori, N., Ohsugi, M. & Yamamoto, T. The Plk1 meiosis I. Nature 441, 53–61 (2006). Proc. Natl Acad. Sci. USA 104, 4383–4388 (2007). target Kizuna stabilizes mitotic centrosomes to ensure 249. Kitajima, T. S. et al. Shugoshin collaborates with 264. Petronczki, M., Glotzer, M., Kraut, N. & Peters, J. M. spindle bipolarity. Nat. Cell Biol. 8, 1095–1101 protein phosphatase 2A to protect cohesin. Nature Polo-like kinase 1 triggers the initiation of cytokinesis (2006). 441, 46–52 (2006). in human cells by promoting recruitment of the 234. De Luca, M., Lavia, P. & Guarguaglini, G. A functional 250. Tang, Z. et al. PP2A is required for centromeric RhoGEF Ect2 to the central spindle. Dev. Cell 12, interplay between Aurora-A, Plk1 and TPX2 at spindle localization of Sgo1 and proper chromosome 713–725 (2007). poles: Plk1 controls centrosomal localization of segregation. Dev. Cell 10, 575–585 (2006). 265. Hutterer, A. et al. Mitotic activation of the kinase Aurora-A and TPX2 spindle association. Cell Cycle 5, 251. Hansen, D. V., Loktev, A. V., Ban, K. H. & Jackson, Aurora-A requires its binding partner bora. Dev. Cell 296–303 (2006). P. K. Plk1 regulates activation of the anaphase 11, 147–157 (2006). 235. Kumagai, A. & Dunphy, W. G. Purification and promoting complex by phosphorylating and 266. Seki, A. et al. Plk1- and β-TrCP-dependent degradation molecular cloning of Plx1, a Cdc25-regulatory kinase triggering SCFβTrCP-dependent destruction of the APC of Bora controls mitotic progression. J. Cell Biol. 181, from Xenopus egg extracts. Science 273, 1377–1380 inhibitor Emi1. Mol. Biol. Cell 15, 5623–5634 65–78 (2008). (1996). (2004). Acknowledgements 236. Roshak, A. K. et al. The human polo-like kinase, PLK, 252. Moshe, Y., Boulaire, J., Pagano, M. & Hershko, A. I regret that I am unable to cite numerous original and signifi- regulates cdc2/cyclin B through phosphorylation and Role of polo-like kinase in the degradation of early cant papers because of space constraints. I thank C. McInnes activation of the cdc25C phosphatase. Cell Signal 12, mitotic inhibitor 1, a regulator of the anaphase for the critical review of the manuscript. I thank T. Berg, 405–411 (2000). promoting complex/cyclosome. Proc. Natl Acad. Sci. S. Kappel, M. Sanhaji and Y. Matthess for assistance in the 237. Watanabe, N. et al. M-phase kinases induce USA 101, 7937–7942 (2004). preparation of the list of references and figures. This work phospho-dependent ubiquitination of somatic 253. Eckerdt, F. & Strebhardt, K. Polo-like kinase 1: was supported by the Deutsche Krebshilfe, the Else–Kröner– Wee1 by SCFβ-TrCP. Proc. Natl Acad. Sci. USA 101, target and regulator of anaphase-promoting complex/ Fresenius–Stiftung, the Carls–Stiftung and LOEWE Centre 4419–4424 (2004). cyclosome-dependent proteolysis. Cancer Res. 66, Frankfurt. 238. Inoue, D. & Sagata, N. The polo-like kinase Plx1 6895–6898 (2006). interacts with and inhibits Myt1 after fertilization of 254. Golan, A., Yudkovsky, Y. & Hershko, A. The cyclin- Competing interests statement Xenopus eggs. EMBO J. 24, 1057–1067 (2005). ubiquitin ligase activity of cyclosome/APC is jointly The author declares no competing financial interests. 239. Jackman, M., Lindon, C., Nigg, E. A. & Pines, J. activated by protein kinases Cdk1-cyclin B and Plk. Active cyclin B1-Cdk1 first appears on centrosomes J. Biol. Chem. 277, 15552–15557 (2002). in prophase. Nature Cell Biol. 5, 143–148 (2003). 255. Kraft, C. et al. Mitotic regulation of the human DATABASES 240. Yuan, J. et al. Cooperative phosphorylation including anaphase-promoting complex by phosphorylation. Entrez Gene: http://www.ncbi.nlm.nih.gov/gene the activity of polo-like kinase 1 regulates the EMBO J. 22, 6598–6609 (2003). Plk1 | PLK1 | Plk2 | PLK2 | Plk3 | PLK3 | Plk4 | PLK4 subcellular localization of cyclin B1. Oncogene 21, 256. Nasmyth, K. Segregating sister genomes: UniProtKB: http://www.uniprot.org 8282–8292 (2002). the molecular biology of chromosome separation. Aurora A | PLK1 | PLK2 | PLK3 | PLK4 | PLK5 | polo 241. Yamashiro, S. et al. Myosin phosphatase-targeting Science 297, 559–565 (2002). FURTHER INFORMATION subunit 1 regulates mitosis by antagonizing polo-like 257. Alexandru, G., Uhlmann, F., Mechtler, K., Author’s homepage: kinase 1. Dev. Cell 14, 787–797 (2008). Poupart, M. A. & Nasmyth, K. Phosphorylation http://www.molgyn.kgu.de/genesilencer/ 242. van Vugt, M. A., Bras, A. & Medema, R. H. Restarting of the cohesin subunit Scc1 by Polo/Cdc5 kinase the cell cycle when the checkpoint comes to a halt. regulates sister chromatid separation in yeast. ALL LinKs Are AcTive in The onLine PDf Cancer Res. 65, 7037–7040 (2005). Cell 105, 459–472 (2001).