In vivo propagation of parasite material was performed in mongolian jirds (Meriones unguiculatus), which were raised and housed at the local animal facility of the Institute of Hygiene and Microbiology, University of Würzburg. This study was performed in strict accordance with German (Deutsches Tierschutzgesetz,TierSchG, version from Dec-9-2010) and European (European directive 2010/63/EU) regulations on the protection of animals. The protocol was approved by the Ethics Committee of the Government of Lower Franconia (Regierung von Unterfranken) under permit number 55.2-2531.01-31/10.

PLK inhibitors BI 2536 [23] and BI 6727 (Volasertib) [27] were purchased from Axon Medchem (Groningen, The Netherlands) and Selleckchem.com (München, Germany), respectively. Both inhibitors were dissolved in dimethyl sulfoxide (DMSO) from Sigma Aldrich (D8418-50ML) as 10 mM stock solutions and were stored at −80°C until use (according to the manufacturer's instructions).

All experiments were performed using the E. multilocularis isolates H95 (cloning procedures; drug treatment) [28] and GH09 (whole mount in situ hybridization) [28]. Mongolian jirds (M. unguiculatus) were used for in vivo propagation of the parasite by intraperitoneal passages as previously described [26]. Co-cultivation of metacestode vesicles with host cells was carried out essentially as previously described [26]. Axenic cultivation of metacestode vesicles was performed as described by Spiliotis et al. [25]. Echinococcus primary cells were isolated and cultivated under axenic conditions as described by Spiliotis et al. [13]. Conditioned medium (A4 medium) was prepared by seeding 1×10⁶ rat Reuber hepatoma cells [26] together with 100 ml DMEM medium (Life technologies) in a culture flask, followed by 1 week incubation. Subsequently, the supernatant was removed and sterile filtrated (A4 medium). Inhibitor tests on metacestode vesicles were performed under axenic culture conditions (nitrogen atmosphere [13]) in A4 medium as previously described by Hemer et al. [29] and Gelmedin et al. [30], with up to ten vesicles per well of a 6-well (5 ml vol. per well) culture plate. Primary cell cultures were set up essentially as described by Spiliotis et al. [24] and the amount of isolated cells was subsequently measured indirectly through densitometry. 1 Unit of primary cells was defined as the amount that yields an OD₆₀₀ of 0.02 in Phosphate Buffered Saline (PBS). 50 Units isolated primary cells (∼15.000 cells) were then seeded in a 48-well plate with 1.3 ml A4-medium. Medium was changed every second day and fresh inhibitor was added. Inhibitors were used in final concentrations of 5, 10, 25, 50, and 100 nM. All experiments were performed with at least three technical and biological replicates. Linear regression analysis was carried out using MicrosoftExcel-2007. Error bars in figures represent standard deviation. Differences were considered significant for p values below 0.05 (indicated by (*)), for p between 0.001 and 0.01 (**), and for p<0.001 (***). For p>0.5, differences were considered non-significant.

Proliferation of stem cells was investigated by staining of newly synthesized DNA using the Click-iT EdU cell proliferation assay (Invitrogen, C10337). Metacestode vesicles from hepatocyte co-culture [26] were washed once with 1× PBS and carefully transferred into a 15 ml tube containing 3 ml A4-medium. A final concentration of 50 µM EdU (Component A) was added and the metacestode vesicles were incubated for 5 h at 37°C (with gentle agitation every 45 min). For fixation, metacestode vesicles were transferred to a Petri-dish and carefully opened with a syringe tip to remove hydatid fluid. After intense washing with 1× PBS, fixative (4% paraformaldehyde (PFA) in 1× PBS) was added and samples were incubated for 1 h at room temperature. The samples were washed again (1× PBS) to remove the fixative, and then transferred to a 1.5 ml tube. The EdU labeling procedure was then carried out according to the manufacturer's instructions (Invitrogen, C10337), but with extended incubation times. Metacestode vesicles were finally analyzed by fluorescence microscopy.

For long-term treatment (21 days) with BI 2536, metacestode vesicles were cultivated in A4-medium for 21 days with inhibitor (medium and inhibitor change every second day), followed by three days recovery without treatment. As a control, medium with and identical amount of DMSO (without inhibitor) was used. Randomly chosen metacestode vesicles (t = 4) were isolated and EdU stained as outlined above.

Of each metacestode vesicle, three randomly chosen sections of the GL were analyzed by microscopy and EdU positive cells were counted. The number of EdU positive cells was calculated to cells per mm² of the GL.

For short-time inhibitor experiments, metacestode vesicles were treated in vitro for 24, 48 and 72 hours with 50 nM or 100 nM BI 2536, followed by 24 h regeneration without inhibitor. EdU staining and calculation of Edu positive cells per mm² of GL were subsequently performed as described above.

The E. multilocularis genome sequence assembly [28] was used as available in GeneDB under http://www.genedb.org/Homepage/Emultilocularis. Published sequences of human Plk1 (GenBank accession number: P53350) and SmPlk1 (AY747306) were employed in extensive BLASTP searches on the genome sequence. Amino acid sequence and domain predictions were carried out using UniProt (http://www.uniprot.org/) and SMART (http://smart.embl-heidelberg.de/) software. Based on the genomic sequence information, primers were designed to amplify the full EmPlk1 coding sequence. Total RNA was isolated from metacestode vesicles with TRIzol reagent and reverse transcribed into cDNA as previously described [30]. The cDNA then served as a template for gene amplification by PCR using primers AKO-053 (5′-GAC TTC TGC CCG GGT ATG GAT A-3′) and AKO-054 (5′-GGA AGA CGG CAA ACA TGT GAT-3′). The PCR product was subsequently cloned into pJET 1.2 (Thermo Scientific CloneJET PCR Cloning Kit) and sequenced.

According to the previously described protocol for heterologous expression of SmPlk1 in Xenopus oocytes [21], mutated versions of EmPlk1 were produced which included a constitutively active form of the kinase (T₁₇₉D), a version that cannot be phosphorylated at T₁₇₉ (T₁₇₉V), and kinase dead versions of wild-type EmPlk1 (wt^KD) and T₁₇₉D (T₁₇₉D^KD). The kinase dead versions were generated by replacing the highly conserved active loop motif D₁₆₃FG for D₁₆₃SV. Kinase domain mutants were generated by mutagenesis PCR using the following primer combinations: T₁₇₉D (AKO-138, 5′-GGT GAA ATG AAG AAG GAC TTA TGT GGG ACG CCA AAC TAT ATT GCT CC-3′ and AKO-139, 5′-CCA CAT AAG TCC TTC TTC ATT TCA CCT TCT TTA GTA ATT CTA G-3′), T₁₇₉V (AKO-140, 5′-GGT GAA ATG AAG AAG GTA TTA TGT GGG ACG CCA AAC TAT ATT GCT CC-3′ and AKO-141, 5′-CCA CAT AAT ACC TTC TTC ATT TCA CCT TCT TTA GTA ATT CTA G-3′), and D₁₆₃SV (AKO-146, 5′-GAC ATG ATT GTA AAG ATC GGG GAT TCG GTG TTG GCC TCT AGA ATT ACT AAA GAA GG-3′ and AKO-147, 5′-GTA ATT CTA GAG GCC AAC ACC GAA TCC CCG ATC TTT ACA ATC ATG TCA TCA TTT AAA AAC AGA TTG GC-3′). For mutagenesis, the emplk1 reading frame was cloned into the pBAD TOPO/Thio expression vector (Life Technologies). Mutations were then introduced, employing the above listed primers, using the QuickChange Site-directed mutagenesis kit (Agilent Technologies) according to the manufacturer's instructions. All plasmid constructs were finally sequenced to verify wild-type sequences and the successful introduction of the desired mutation. Wild-type and mutant versions of the EmPlk1 reading frame were further cloned into the pSecTag2/Hygro A vector (Life Technologies) that contains a T7 promoter sequence. cRNA encoding EmPlk1 proteins was synthesized in vitro using the T7 mMessage mMachine Kit (Ambion, USA) and injected in stage VI Xenopus laevis oocytes according to [31].

Heterologous expression of EmPlk1 wild-type and mutant forms in Xenopus oocytes, in vitro treatment of oocytes with the inhibitors BI2536 and BI6727, and germinal vesicle breakdown assays were conducted essentially as previously described by Long et al. [21] for SmPlk1. As a positive control for GVBD, oocytes were stimulated with progesterone (PG), the natural inducer. All experiments were carried out on samples composed of 20 oocytes originated from three different Xenopus females.

Total RNA was isolated using TRIzol as previously described [32] from 2, 5, and 11 day old primary cell cultures, from dormant and pepsin/low pH activated protoscoleces [33], and from mature metacestode vesicles. Isolated RNA was DNase treated (RQ1 RNase-free DNase, Promega) followed by phenol/chloroform extraction. RNA concentration was quantified by spectrophotometry and 750 ng of each stage were used for reverse transcription (RT) as previously described [32]. Intron-flanking, gene specific primers for emplk1 (AKO-57; 5′-GAG CAT GTT CAG TGT GAT GG-3′ and AKO-60; 5′-CGA TCT ATC ATA TCG TAG GCG-3′) were used to estimate the semi-quantitative expression profile by PCR employing a protocol of 30 cycles of 30 sec at 94°C, 30 sec at 58°C, and 1 min at 72°C. The constitutively expressed gene elp [34] served as a control.

Metacestode vesicles of isolate GHO9 with developing protoscoleces were used for in situ hybridization according to Koziol et al. [14]. A 1.3 kb fragment of the emplk1 open reading frame was amplified and subcloned into pDrive (Qiagen) using primers AKO-55 (5′-CTC TCA TGG AAC TGC ATA AGA G-3′) and AKO-60 (5′-CGA TCT ATC ATA TCG ATG GCG-3′). Digoxygenin-labelled antisense and sense RNA probes were in vitro transcribed using the T7 or SP6 promoter of the linearized vector as previously described [14]. Detection was performed using anti-digoxygenin antibodies coupled to alkaline phosphatase, and NBT/BCIP as colorimetric substrates, as described [14].

The complete emplk1 cDNA sequence reported in this paper was submitted to the GenBank database and is available under accession number HG931729.

By BLASTP genome mining of the available E. multilocularis genome sequence [28] (http://www.genedb.org/Homepage/Emultilocularis) using the full length sequences of human Plk1 [35] and S. mansoni SmPlk1 [21] as queries, one single locus encoding an Echinococcus Plk1 ortholog (EmuJ_000471700) was identified on chromosome 3. Due to its homologies (see below), the respective gene was designated emplk1 (E. multilocularis Polo-like kinase 1) encoding the protein EmPlk1. The emplk1 gene spanned a genomic region of 3.174 bp and, like the SmPlk1 encoding gene of S. mansoni [21], comprised 7 exons, separated by 6 introns. All introns displayed canonical GT-AG dinucleotide sequences at the 5′ splice donor and the 3′ splice acceptor sites. Informed by the emplk1 genomic sequence, primers were designed to PCR-amplify the entire emplk1 reading frame from metacestode cDNA preparations. The cDNA fragment was cloned and fully sequenced, which confirmed the exonic gene sequence as determined by the E. multilocularis genome project. The full length EmPlk1 reading frame comprised 1833 bp and coded for a protein of 610 amino acids, EmPlk1, with a calculated molecular mass of 69.5 kDa. Analysis of the EmPlk1 protein sequence by SMART [36] identified an N-terminal STK catalytical domain (between Y₂₂ and F₂₇₄) and a C-terminal protein binding domain, which included two Polo box domains (Figure 1). The EmPlk1 kinase domain displayed sequence motifs characteristic of all 11 typical sub-domains (I–XI) of protein kinases and included all residues previously determined to be invariable for protein kinases [37] at the respective positions (Figure 1). Particularly the sequence motifs HRDLKxxN (sub-domain VI) and GTPNYIAPE (VIII) are strong indicators for STK activity [37]. In human Plk1, residue T₂₁₀ was previously shown to be the major phosphorylation site of mitotic PLKs [15] and a respective threonine residue was also conserved in schistosome SmPlk1 (T₁₈₂) [21]. In EmPlk1, this residue was also conserved (T₁₇₉; Figure 1). As in SmPlk1 [21] and human Plk1 [35], the ATP binding site of EmPlk1 contained a GxGGFAxC motif (Figure 1), which is a PLK-typical variation of the canonical GxGxxGxV motif found in the majority of protein kinases [37]. Overall, the EmPlk1 kinase domain displayed significant homologies to the kinase domains of well investigated PLKs such as S. mansoni SmPlk1 (68% identical residues), human Plk1 (62%), Drosophila Polo (60%), and Xenopus Plx1 (60%).

Protein-protein interaction and cellular localization of PLKs is regulated by the C-terminally located protein binding domain [15]. As typical for Plk1-like PLKs, the EmPlk1 protein binding domain contains two Polo boxes (Y₃₇₈-T₄₄₁ and W₄₇₅-Y₅₄₅), which are separated from the kinase domain by a non-conserved linker region. In human Plk1, three residues (W₄₁₄, H₅₃₈, K₅₄₀) have previously been shown to be essential for the binding to phospho-S/T binding motifs, and all three residues are also perfectly conserved in EmPlk1 (W₃₇₁, H₄₉₉, K₅₀₁; Figure 1).

Kothe et al. [38] previously identified several amino acid residues that are important for binding of the PLK inhibitor BI 2536 to human Plk1. In particular, the absence of a bulky side chain at position 132 (L₁₃₂ in human Plk1) was an important specificity determinant that ensured optimal binding of BI 2536 to the Plk1 subfamily of PLKs [38]. In EmPlk1, all these residues, including the leucine residue (L₁₀₁), were conserved at the respective positions (Figure 1).

Taken together, all above analyses clearly identified EmPlk1 as a member of the Plk1 subfamily of PLKs, indicated that the Echinococcus protein is most probably enzymatically active, and that the PLK inhibitor BI 2536 should be able to bind to the parasite-derived kinase.

To determine whether emplk1 is expressed in Echinococcus larval stages relevant to the infection of the intermediate host, semi-quantitative RT-PCR experiments were carried out. As relevant parasite stages, we chose mature metacestode vesicles and protoscoleces before and after activation by pepsin/low pH (mimics transition into the definitive host), to cover late stages of the infection. In our established primary cell cultivation system [13], [24], E.multilocularis germinative cells are capable of developing into metacestode vesicles similar to the oncosphere metacestode transition process [7]. Hence, in order to cover early stages of the infection, we also included primary cell cultures at different time points (2, 5, 11 days) of development. Total RNA was isolated and, after cDNA preparation, emplk1 gene specific PCR was carried out on serial dilutions. As shown in Figure 2, emplk1 transcripts could be clearly identified in all larval stages tested, and particularly prominent expression was observed in primary cell cultures, which typically contain large proportions of parasite stem cells [13], [14].

During the E. multilocularis genome project, preliminary deep sequencing transcriptome data were generated for primary cell cultures (2 and 11 days old) as well as for metacestode vesicles and activated/dormant protoscoleces [28]. When we analyzed these profiles we found particularly high expression in primary cell cultures after 2 days of cultivation, which was reduced in primary cells after 11 days, and basal in metacestode vesicles and protoscoleces (Figure S1). This verified the RT-PCR data mentioned above and, again, indicated possible stem cell specific expression of emplk1 since young primary cell cultures contain particularly high percentages of germinative cells, which steadily decline during development (differentiation) into mature vesicles [14].

We recently investigated cellular proliferation profiles in E. multilocularis development [14] and demonstrated that mitotically active parasite stem cells are distributed throughout the germinal layer, and are strongly accumulated in brood capsule and protoscolex buds. In late stage protoscoleces, stem cells are prominently located at the base of developing suckers, but are also present in the posterior body [14]. To investigate a possible stem cell specific expression of emplk1, we therefore carried out in situ hybridization experiments using a recently established protocol that is applicable to metacestode vesicles [14]. As depicted in Figure 3, prominent emplk1 signals were obtained for all regions of parasite larvae that typically contain large numbers of proliferating stem cells, such as early brood capsules (Figure 3D), and developing protoscoleces (Figure 3E,F). Furthermore, germinal layer cells with a distribution highly reminiscent of germinative stem cells stained positive for emplk1 (Figure 3C). Taken together, these data strongly indicated that emplk1 is specifically expressed in parasite stem cells.

For functional studies on the schistosome PLK SmPlk1, the heterologous Xenopus oocyte expression system has previously been employed [21], and in the present study we used this system to further characterize EmPlk1. To this end, we first generated a mutant form of EmPlk1 in which T₁₇₉ of the activation loop was replaced by phospho-mimetic aspartate (T₁₇₉D), thus yielding a constitutively active form of the enzyme (similar to [21]). We also produced a mutant in which T₁₇₉ was replaced by valine (T₁₇₉V), which would prevent phosphorylation, and thus activation, at T₁₇₉ in Xenopus oocytes. Finally, we produced ‘kinase dead’ versions by replacing the D₁₆₃FG motif of the active loop by D₁₆₃SV in both the wild-type and the T₁₇₉D background (wt^KD; T₁₇₉D^KD). In Xenopus oocytes, it has already been demonstrated that the injection of mRNA encoding activated forms of Plx1 (the Xenopus Plk1-like kinase) or SmPlk1 can induce meiosis resumption, which results in germinal-vesicle breakdown (GVBD) [21]. We therefore injected mRNAs encoding the wild-type and mutant forms of EmPlk1 into Xenopus oocytes and carried out GVBD assays.

As shown in Table 1, injection of (non-activated) wild-type EmPlk1 alone was ineffective in inducing GVBD. However, injection of the (activated) T₁₇₉D mutant led to GVBD in 90% of oocytes, which is comparable to activities previously determined for SmPlk1 in this system [21]. The induction of GVBD by an activated mutant of SmPlk1 in Xenopus oocytes has previously been shown to involve increased phosphorylation of Xenopus Cdc25C [21] and, although we did not specifically test this for EmPlk1, we assume that the Echinococcus enzyme is also able to directly activate the Xenopus downstream target. In the case of the kinase dead version of activated EmPlk1 (T₁₇₉D^KD), on the other hand, only a basal rate (10%) of GVBD was observed (Table 1), indicating that the kinase activity of EmPlk1 is responsible for meiosis resumption in Xenopus. The T₁₇₉V version of EmPlk1 was also completely inactive in inducing Xenopus oocyte GVBD, indicating that phosphorylation of EmPlk1 at T₁₇₉ is essential for enzymatic activity. It should be noted that BI 2536 was rather ineffective in inhibiting GVBD in response to the positive control progesterone (Table 1). This is in line with previous investigations showing that progesterone-induced GVBD of Xenopus oocytes occurs via several pathways, among which are the cAMP pathway (involving protein kinase A) and the mitogen-activated protein kinase (MAPK) pathway (including Mos, an oocyte-specific MAPKKK) [39]. Furthermore, meiosis entry of Xenopus oocytes in response to progesterone also depends on a balance between Cyclin B synthesis and the activity of Myt1, a member of the Wee1 family of inhibitory kinases, in a Plx1- and MAPK-independent manner [40]. Hence, although Plx1 is required for the activation of Cdc25C in Xenopus oocytes [41], leading to accelerated meiosis resumption, Plx1 inhibition can eventually not prevent GVBD. The fact that even high concentrations of BI 2536 do not significantly affect progesterone-induced GVBD in our experiments thus indicates that this compound is not generally cytotoxic to Xenopus oocytes and that the BI 2536 effects on EmPlk1 T₁₇₉D-induced GVBD solely result from the specific inhibition of the parasite enzyme.

Taken together, these experiments verified that EmPlk1 is an enzymatically active kinase that can stimulate meiosis resumption in the Xenopus oocyte system similarly to SmPlk1.

Having established that EmPlk1 can induce GVBD in Xenopus oocytes, we then tested whether an available inhibitor, BI 2536, originally designed against human Plk1 [23], is also able to affect the parasite protein. As shown in Table 1, already at concentrations as low as 5 nM, the capability of EmPlk1 to induce GVBD in Xenopus oocytes started to diminish, and was completely abolished at concentrations of 50 nM or higher. Similar results were observed when we used BI 6727, another Plk1-specific inhibitor with improved pharmacokinetic profile [27], although in this case slightly higher concentrations were necessary to induce EmPlk1 inhibition (data not shown). Taken together, these experiments indicated that EmPlk1 can be inhibited by BI 2536 and BI 6727 in a concentration dependent manner.

Having established that the PLK inhibitor BI 2536 affects the enzymatic activity of EmPlk1, we next investigated whether this compound could also inhibit the formation of metacestode vesicles from parasite stem cells. To this end, we set up Echinococcus primary cell cultures and measured the formation of mature metacestode vesicles in the presence of different concentrations of BI 2536. As shown in Figure 4, metacestode formation was already significantly reduced in the presence of 5 nM BI 2536, while concentrations of 25 nM or higher almost completely prevented parasite development. We also tested BI 6727 in the primary cell culture system and, as BI 2536, this inhibitor prevented metacestode vesicle formation in a concentration dependent manner, although at slightly higher concentrations than BI 2536 (data not shown). Since the formation of metacestode vesicles in the primary cell system crucially depends on proliferating stem cells [14], and since we had already shown that EmPlk1 is specifically expressed in this cell type (see above), we therefore concluded that the inhibition of EmPlk1 by BI 2536 and BI 6727 either resulted in stem cell killing, or at least prevented stem cell proliferation, in the primary cell system.

We then tested the effects of BI 2536 on mature metacestode vesicles, which is the actual target stage of chemotherapy in AE [1], [5]. Interestingly, as shown in Figure S2, even incubation for 21 days in the presence of 100 nM BI 2536 did not lead to structural disintegration or collapse of metacestode vesicles which, however, had slightly reduced sizes and were no longer capable of growing. This was similar to an approach we had recently undertaken concerning metacestode treatment with the ribonucleotide reductase inhibitor hydroxyurea (HU), which is toxic to cells that undergo proliferation [14]. Although HU treatment specifically eliminated germinative cells in metacestode vesicles, these remained structurally intact for several weeks, indicating that this parasite stage is able to survive for long periods under conditions of slow cellular turnover [14]. We therefore tested whether BI 2536 treatment of metacestode vesicles might specifically eliminate the germinative cell population and carried out EdU incorporation experiments. To this end, metacestode vesicles were incubated for 21 days in presence of different concentrations of BI 2536. After recovery for 3 days without inhibitor, proliferating cells were detected by EdU pulse labeling [14]. As shown in Figure 5, control vesicles displayed a typical pattern of proliferating germinative cells within the germinal layer. In samples treated with 10 nM BI 2536, however, the number of proliferating stem cells was reduced by ∼50%. A statistically significant reduction to ∼10% of normal numbers was observed after incubation with 25 nM BI 2536 and in the presence of higher inhibitor concentrations (50, 100 nM), cell proliferation in the germinal layer was completely abolished (Figure 5). Like in our previous experiments using HU as an inhibitor [14], no effects of BI 2536 treatment were observed on differentiated cells. As expected from the absence of proliferating cells, even after 3 weeks of further cultivation, these vesicles did not resume growth or proliferation capacity (data not shown). Using vesicles after 21 day treatment with 50 or 100 nM BI 2536 as a source for parasite primary cell cultivation, we also never obtained cultures that formed typical cell aggregates [13], [14] or mature metacestode vesicles (data not shown). Finally, we also carried out short term treatment (24, 48, 72 h) of metacestode vesicles with 50 nM and 100 nM BI 2536 concentrations, followed by 1 day recovery and 5 h EdU pulses. As shown in Figure 6, already after 24 h in the presence of 50 nM BI 2536, the number of proliferating cells within the germinal layer was reduced to ∼17% of the normal proliferating cell number, and it was further diminished after incubation for 72 h or at higher concentrations (100 nM; Figure 6).

Taken together, the experiments outlined above clearly indicated that BI 2536 specifically targeted the germinative stem cell population of metacestode vesicles and either led to stem cell killing or long-term mitotic arrest.