Background
Malignant melanoma is the deadliest form of skin cancer, in addition to having one of the most increasing incidence rates of all cancer forms [
1]. Although curable by surgical excision at an early stage, patients diagnosed with metastatic melanoma (stage IV) have had a median survival of 6–10 months [
2]. Despite recent year’s advances leading to new treatment options such as Ipilimumab, Vemurafenib, Trametinib and Dabrafenib, there are still no curative treatment alternatives for the majority of the patients with advanced disease [
3,
4].
In response to DNA-damage, the dividing cell is arrested through activation of checkpoint mechanisms in order to allow time for DNA-repair to be completed. If the damage is too severe, apoptosis or senescence is induced to ensure that unrepaired DNA-damage is not passed on to future generations of cells. Loss of checkpoint mechanisms is frequently found in cancer, a trait which can be exploited in cancer therapy. Wee1 is a kinase involved in checkpoint regulation that in response to DNA-damage or replication stress can halt the cell cycle progression in S- and G2 phases by adding inhibitory phosphorylations (Tyr15) on cyclin-dependent kinases CDK2 and CDK1, respectively [
5]. In our previous study we found that Wee1 was up-regulated in human melanomas as compared to benign nevi, and that high expression of Wee1 was associated with poor disease-free survival [
6]. Likewise, over-expression of Wee1 (protein and/or mRNA) has been reported in osteosarcoma, glioblastoma and ovarian- and vulvar squamous cell carcinomas, thus emphasizing its potential as a therapeutic target in cancer [
6‐
10]. Furthermore, targeting Wee1, either by siRNA mediated silencing or inhibitors, has in several studies, including ours, been shown to lead to increased DNA-damage and apoptosis [
6,
11,
12].
Although mono-targeting of Wee1 has shown anti-tumour effect in some cancer cell lines, a stronger effect has been observed when combining Wee1 inhibitors with for instance DNA-damaging agents, Heath Shock Protein 90 inhibitors and more recently inhibitors of other cell cycle regulatory proteins such as Chk1/2 [
13‐
16]. The Chk1/2 kinases are key regulators of DNA-damage surveillance pathways and DNA repair. Chk2 is a protein that is stably expressed throughout the cell cycle and is activated in response to DNA-damage; in particular DNA double-strand breaks through the ATM-Chk2-p53-p21 pathway. The expression of Chk1, on the other hand, is primarily found in S- and G2 phases, where it is active even in the unperturbed cell cycle. During normal cell cycle progression, Chk1 has been shown to regulate replication forks during DNA replication and entry into mitosis. Chk1 activation in response to DNA-damage is preferentially triggered through the ATR-Chk1-CDC25 pathway, although some cross-talk between Chk1 and ATM has been reported (reviewed in [
17]. Chk1 has been shown to activate Wee1 in Xenopus extracts and yeast; however such a relationship has not been recapitulated in higher eukaryotes [
18,
19]. Previous studies have shown an advantageous effect of combining Wee1 and Chk1/2 inhibitors as compared to mono-targeting in a variety of cancer cell lines, and points to complementary functions of the kinases [
13,
20,
21].
In the present study we investigated the combined use of the Wee1 inhibitor MK1775 with a Chk1/2 inhibitor (AZD7762) in a panel of metastatic melanoma cell lines. While both inhibitors had an effect as mono-agents, combined administration gave a stronger anti-tumour effect both in vitro and in xenografts models. Co-treatment led to increased dephoshorylation of CDK1, DNA-damage, premature mitosis and apoptosis. In summary, our results warrant further evaluation of combined use of Wee1 and Chk1/2 inhibition in malignant melanoma.
Methods
Cell lines and growth conditions
The human metastatic melanoma cell lines WM239, WM45.1, WM983B and WM1366 were kindly provided by Prof. Meenhard Herlyn (the Wistar institute, Philadelphia, USA) [
22,
23]. The FEMX-1 cell line was established at the Radium hospital [
24]. The ‘Patient 3’ cell line was a kind gift from Prof. Peter Hersey (Royal North Shore Hospital, Sydney, Australia) [
25]. All cell lines were maintained in RPMI-1640 medium (LONZA, Verviers, Belgium) supplemented with 5 % Fetal Calf Serum (Biochrom, KG, Berlin, Germany) and 2 mM L-glutamine (LONZA, Verviers, Belgium). The cells were grown in culture at 37 °C in humidified conditions containing 5 % CO
2, either as monolayer cultures in 75 cm2 bottles or in 96 flat-bottom well plates. Normal human melanocytes (FOMA4) and fibroblasts (FF144sc) were isolated from human foreskin and cultured in 254CF (Invitrogen corporation, CA, USA) and DMEM 10 % FBS medium, respectively, as previously described [
6].
Spheroids were generated by plating suspended cells (500–4000 cells/well, dependent on the cell line) in Corning® 96 Well Clear Round Bottom Ultra Low Attachment Microplates (Corning, MA, USA). Spheroid formation was allowed for 3 days prior to treatment. Images of spheroids were obtained using phase-contrast on an Olympus IX81 microscope with a 4× objective. Spheroid volume was calculated using Olympus Soft Imaging Solution Gm6H software. A minimum of two independent biological experiments were conducted, where each experiment contained at least four parallels of the individual treatment options.
Chemical inhibitors
Wee1 inhibitor MK1775 and Chk1/2 inhibitor AZD7762 were purchased from Selleck Chemicals (TX, USA) and used for time intervals and concentration indicated in the text.
Small interfering RNA (siRNA) transfection
All cell lines were plated in either 6-well plates (1.5 × 105 cells/well) or in 96-well plates (5 × 103 cells/well) 24 h in advance of the transfection. The cells were transfected with 10nM siRNA targeting Wee1 (OligioID; ‘VHS50841’), Chk1 (OligioID: ‘VHS40226’) or RNAi negative control duplexes (Negative Control LOW GC, 12935–200) using Lipofectamine™ RNAiMAX transfection reagents (all reagents from Invitrogen corporation, CA, USA). Transfection was allowed for 5 h before the medium was replaced with RPMI w/5 % FCS and 2 mM L-glutamine. Transfected cells were harvested after 48 h for further analysis.
Viability assays
Four thousand cells per well were seeded in 96-well plates and left to attach overnight, before treatment with MK1775 and/or AZD7762 for 48 h. The growth inhibitory effects of mono- and combined treatments were measured using the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) (MTS) assay (Promega, WI, USA). Absorbance was measured at 490 nm using ASYS UVM340 96-well plate reader. Alternatively, viability was assessed using the CellTiter-Glo® Luminescent Cell Viability Assay kit (Promega) following the manufacturer's protocol. Luminescence was measured using GloMax® Luminometer (Promega). Viability of treated cells was normalized to the untreated control cells. Each experiment was performed with three parallel observations and repeated at least three times.
Calcusyn analysis
Synergy was determined by the Chou and Talalay Combination Index (C.I.) [
26] for non-exclusive treatments (treatments affecting different targets or sites of the same target), and calculated by Calcusyn software (BioSoft, Feruson, MO, USA). Of note, this method requires that a dose effect curve for each drug is made, in which the data-points give a good r-value (>0.90 for cell systems) [
27]. Given the variation in dose effect of the drugs in the different cell lines, the concentrations of the inhibitors were adjusted for the individual cell lines (Additional file
1: Figure S1) in order to abide to the requirements of the method.
Western blot analysis
Cells were harvested and western blot conducted as previously described [
6]. Caspase 3 (#9662/#9664 (even mix)), Caspase 8 (#9746), Caspase 9 (#9502), PARP (#9532) and Wee1 (#4936S) primary antibodies were purchased from Cell Signaling (Beverly, MA). α-tubulin (DMIB) was acquired from Calbiochem (Nottingham, UK), whereas Cyclin A (sc-751) and p53 (sc-126) antibodies were obtained from Santa Cruz (Santa Cruz, CA). γ-H2AX (#05-636) and pCDK1
Tyr15 (ab47594) antibodies were acquired from Millipore and Abcam (Cambridge, England), respectively. Following primary hybridization, membranes were washed 3 × 10 min in TBST and hybridized with an appropriate secondary antibody (HPR-conjugated anti-rabbit or anti-mouse IgG antibodies (Promega)) for 1 h at room temperature, with gentle agitation. Membranes were then washed in TBST for 3 × 10 min before being incubated in ECL-plus (GE Healthcare, Chalfont St Gils, UK) for 5 min. Protein bands were visualized by exposing the membranes to X-ray films.
Flow cytometry and barcoding
Cells were harvested by trypzination and fixated in 70 % ice-cold ethanol at −20 °C for a minimum of 24 h. In order to eliminate variation in antibody staining, control as well as treated cells were labelled with different concentrations of Pacific Blue (0.125, 0.031, 0.0062 and 0.00078 ng/μL, respectively) for 30 min at RT in the dark, in accordance with the fluorescent cell barcoding (FCB) technique [
28]. All samples were subsequently mixed in one tube and incubated with primary antibodies (mouse anti-γH2AX (1:500, 05–636, Millipore) and rabbit anti-phospho Histone H3 (ser10) (1:500, CS#9701, Cell Signaling)) diluted in detergent buffer (0.1 % Igepal CA-630, 6.5 mM Na
2HPO
4, 1.5 mM KH
2PO
4, 2.7 mM KCl, 137 mM NaCl, 0.5 mM EDTA [pH 7.5]) containing 4 % nonfat milk for 1 h at RT, and afterwards with secondary antibodies; anti-rabbit Alexa488 and anti-mouse Dyelight549 (1:1000, Invitrogen) for 30 min at RT in the dark. Cells were finally incubated with PBS containing Cell Cycle 633/Fx cycle™ far red stain (1 μL/mL,) and PureLink RNAse A (5 μL/mL) (Invitrogen) for 30 min at 4 °C in the dark. Flow cytometric analysis was performed using a LSRII flow cytometer (BD Biosciences) with Diva software, and the four samples were gated based on the Pacific Blue signal before analyzes of cell cycle distribution, γH2AX and phospho Histone H3 (ser10) expression.
In vivo studies
‘Patient 3’ cells (2 × 106) were subcutaneously injected on each side of the dorsa of nude female mice (athymic nude foxn1 nu) and tumour bearing mice were subsequently divided into groups of 7 mice each. MK1775 and AZD7762 were dissolved and administered as recommended by the manufacturer (Selleck Chemicals). Briefly, 30 mg/kg MK1775 were given orally whereas 25 mg/kg AZD7762 were administered intravenously both as single agents and in combination. The treatments were given 3 times a week for 2 weeks. Control mice received no treatment as we previously have observed no effect of the vehicles alone. Tumour sizes were measured regularly using a calliper, and the volume V was calculated as follows: V = W2 × L × 0.5 (where W and L are tumour width and length, respectively). The experimental protocol was evaluated and approved by the National Animal Research Authority and conducted in accordance with regulations of the European Laboratory Animals Science Association.
Statistical analysis
All statistical analyzes were conducted using SPPS PASWStatistics version 18. Comparison of tumour volume at day 48 was performed with one-way between-groups ANOVA. Post-hoc comparisons were performed using the Tukey HSD tests.
Discussion
In our previous study we showed that high expression of Wee1 is associated with poor disease-free survival in patients with melanoma and that in vitro targeting of the kinase leads to decreased cell viability. To further evaluate Wee1 as a potential target in melanoma, we have in the present study combined a Wee1 inhibitor (MK1775) with a Chk1/2 inhibitor (AZD7762). Our hypothesis is that simultaneous inhibition of two cell cycle control proteins will introduce a high degree of DNA-damage incompatible with cell viability. The combined inhibition led to an increased anti-tumour effect both in vitro and in xenografts models. Furthermore, our in vitro studies showed that co-treatment synergistically decreased viability and increased apoptosis, which may be explained by DNA-damage- and premature mitosis of S-phase cells. Our results provide a rationale for further testing of the treatment regimen in melanoma models.
Mono-targeting of Wee1, either by siRNA mediated knockdown or inhibitors, has shown a potent anti-tumour effect in some cancer cell lines [
6,
35] but only limited effect in other studies [
7,
8]. In line with this, the efficacy of mono-targeting of Wee1 by MK1775 varied in the tested melanoma cell lines in the present study, where low concentrations (200nM) had little effect in some of the cell lines, such as WM1366. Interestingly, a stronger anti-tumour efficacy was observed in the same cell line following exposure to MK1775 in a recent study by Haarberg
et al. [
36], however the experimental conditions in our studies differ with regard to exposure time to the drug and assay for detecting viability.
For the purpose of increasing the efficacy, inhibition of Wee1 (MK1775) has been combined with other therapeutic agents in clinical trials, such as gemcitabine and cisplatin [ClinicalTrials.gov]. Simultaneous inhibition of Wee1 and Chk1, on the other hand, has only been investigated in preclinical studies, but has been found as an eligible combination following siRNA high-throughput screens of acute myeloid leukemia, lung-, prostate- and ovarian cancer [
13,
20,
21]. Combined inhibition of Wee1 and Chk1 using commercial inhibitors have been shown to synergistically enhance the therapeutic efficacy in several cancer cell lines, including the A2058 melanoma cell line [
13,
16,
20,
37]. In agreement with these reports, combined targeting of Wee1 and Chk1/2 led to a synergistic reduction of viability in our cohort of melanoma cell lines. In addition, the same trends were observed for combined siRNA mediated knockdown of Wee1 and Chk1 compared to mono-targeting of either protein. However, siRNA mediated down-regulation and use of small molecular inhibitors are not always comparable [
21]. The experiments using siRNA only targeted Chk1, whereas AZD7762 inhibits both Chk1 and Chk2. Still, a previous study by McNeeley
et al. showed that the anti-tumour activity of AZD7762 following DNA damage is likely related to Chk1- rather than Chk2 inhibition [
38].
Similar to what was observed in monolayer cultures, an increased effect of combined targeting was also found in multicellular spheroids. Likewise, a weak increase in efficacy was observed in melanoma xenografts following combined treatment versus either mono-treatment alone. A previous study showed that combinational inhibition of Wee1 (MK1775) and Chk1 (MK8776) led to increased
in vivo efficacy in neuroblastoma xenografts [
39], providing additional support for our results.
Studies have indicated that the response to Chk1 or Wee1 inhibitors is dependent on the p53 mutational status of the cells, in particular when combined with DNA-damaging agents, while others have reported no such effect [
40,
41]. Partly supporting the first hypothesis, the anti-tumour effect was most pronounced in the p53 mutated WM983B and WM45.1 cells, and less in the p53WT cell line WM239. Although disabling point mutations in
TP53 are only found in 10 % of melanomas, inactivation of the protein is found in approximately 90 % of the tumours (reviewed in [
42]), suggesting that the effect of checkpoint inhibition in melanoma is not strictly dependent on the p53 mutation status. Furthermore, the level of genetic instability has been suggested as a predictor of the response to Wee1 and Chk1/2 inhibition [
43,
44]. In this regard, melanomas have been shown to be one of the most genetically unstable tumour forms [
45,
46]. Combined inhibition of Wee1 and Chk1/2 in primary cultures of melanocytes or fibroblasts had less effect on viability than observed in the melanomas. In accordance with these results, Carrassa
et al. found no synergistic effects of combining Wee1 (MK1775) and Chk1 inhibitors PF-00477736 (or allegedly AZD7762) in the normal fetal human lung cell line MRC-5 [
20]. While both Wee1 and Chk1 are known to have functions in the unperturbed cell cycle, normal cells are less proliferative and have a lower degree of genetic alterations- and instability than tumour cells, which may explain the difference in response to kinase inhibition. However, phase I clinical trials with AZD7762 have been terminated due to cardiac toxicity in patients [
47,
48], and also clinical trials with other Chk1 and Chk1/2 inhibitors (reviewed in [
49]) have been terminated. Concerns have been raised on whether targeting Chk1 may lead to toxicity in normal cells due the multiplicity of functions of the protein in the unperturbed cell cycle (Reviewed in [
50]). Furthermore, embryonic lethality is associated with Chk1 depletion in mice [
51]. On the other hand, the observed toxicities in patients following treatment with various Chk1 inhibitors have appeared as drug-specific rather than class-specific and may thus be a consequence of off-target effects (reviewed in [
49]). Clinical studies with novel Chk1 (MK-8876/SCH 900776) and Chk1/2 (LY2606368) inhibitors with improved specificity are currently ongoing or recruiting (ClinicalTrials.gov).
When further investigating the effect of combined inhibition of Wee1 and Chk1/2, we found a stronger reduction in pCDK1 (tyr15) levels compared to treatment with either drug alone. In line with this, Wee1 has been shown to directly phosphorylate CDKs, and removal of Wee1 is thus expected to reduce such phosphorylations [
52]. Chk1, on the other hand, phosphorylates the CDC25 phosphatases, leading to the latter proteins sequestration or degradation, thereby hindering activating dephosphorylation of pCDK1/2 (tyr15) [
53]. CDK1/2 phosphorylation is thus indirectly maintained by Chk1, and removal of the latter may promote further reduction of pCDK1/2. Inhibition of Wee1 and Chk1 has previously been shown by Davis
et al. to be accompanied by decreased expression of pCDK1 and a synergistic reduction in cell viability [
13]. They further showed that the antiproliferative effect of the inhibitors was partially reversed after CDK inhibition (Roscovitine), suggesting that CDK hyperactivity may be a contributing factor, rather than the sole cause of the antiproliferative effect of Chk1 and Wee1 inhibition. Deregulated activity of CDKs, as a consequence of Wee1 depletion, has in a previous study been shown to induce replication stress and loss of genomic integrity through subsequent nucleotide-shortage and increased firing of replication origins [
54]. In line with this, we observed that inhibition of Wee1 led to increased levels of DNA-damage and apoptosis, as assessed by expression of γ-H2A.X and cleavage of caspase 3 and PARP, respectively. The effect on apoptosis was amplified by combining the inhibitors, supporting a synergistic effect on viability. A further increase in DNA-damage following combined targeting of the kinases was seen in WM239 cells. Similarly, combined inhibition of Wee1 and Chk1 has in a previous study been demonstrated to induce a more intense and durable DNA-damage and anti-tumour effect than either drug alone [
37]. Interestingly, treatment with the Chk1/2 inhibitor or Chk1 siRNA reduced the total protein level of CDK1 in the WM983B cell line. In agreement with this, transcriptional reduction of cell-cycle regulators, such as CDK1, has been observed after Chk1 depletion in somatic cells [
55].
Similar to what has been reported by others [
37,
56,
57], inhibition of Wee1 as well as Chk1 was shown to induce DNA-damage in the S phase of the cell cycle. These findings are in accordance with the proposed function of the kinases in restraining CDK activity during DNA replication (reviewed in [
58]). Furthermore, it has been suggested that the cytotoxic effect of Chk1 inhibition in melanoma cells is most likely due to inhibition of Chk1 in S phase, which drives cells prematurely from late S phase into an aberrant mitosis [
43]. In line with this, combinational treatment with Chk1/2 and Wee1 inhibitors increased the population of S-phase cells co-expressing the mitotic marker, pHH3. These are cells that have entered mitosis without having completed replication, most likely due to the compromised G2/M checkpoint in the absence of Wee1 and/or Chk1. Likewise, an increased proportion of mitotic cells was observed following mono-targeting, and to a further extent by the combination of the inhibitors. Most likely the G2 cells have abrogated the G2/M checkpoint and entered mitosis given the short timeframe of this experiment (3 h).
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
GIM contributed to design, acquiring and interpretation of data as well as drafting of the manuscript. EE, and RF aquired the data. KGF performed the animal experiment, BE contributed to design and interpretation of animal experiments, VN-K contributed to conception, design and interpretation of flow cytometry analyses. AA and VAF contributed to design, interpretation of data and manuscript revision. All authors have read and approved the final manuscript.