Background
Palytoxin (PTX), a toxin isolated from the marine Cnidaria
Palythoa caribaeorum, has a molecular weight of 3300 dalton and was first isolated by Moore and Scheuer[
1]. Recently, it was demonstrated that also Dinoflagellates of the
Ostreopsis genera produce this compound and analogues[
2]. PTX is one of the most toxic non-peptidic natural products known to date. From a chemical perspective, it is a large, very complex molecule with a long polyhydroxylated and partially unsaturated aliphatic backbone, containing 64 chiral centers[
3]. In contrast to most cytotoxins, PTX exerts its activity extracellularly by altering ion equilibria in biological systems[
4]. PTX displays an extraordinary level of cytotoxic activity on a variety of cell lines and it develops a wide spectrum of pharmacological effects such as cellular disruption, binding of the toxin to its receptor[
5], and modulation of protein kinase signalling cascades[
6]. Other studies highlight the cytoskeleton as an early target for the toxic effects of PTX and its analog ostreocin-D on intestinal[
7] and neuroblastoma cells[
8].
Most studies focused on the function and mechanism of PTX which acts through the Na
+, K
+-ATPase[
9], H
+, K
+ –ATPase[
10], interaction with ion channels, and binding reaction to the Na/K pump[
5,
11,
12]. PTX targets the Na
+, K
+ ATPase via binding and locking it in a position allowing passive transport of both the sodium and potassium ions, thereby destroying the ion gradient that is essential for most cells[
13]. The Na
+/K
+-transporting ATPase subunit alpha-1 is an enzyme that in humans is encoded by the ATP1AL1 gene[
14]. Dysfunctions in the Na
+, K
+-ATPase pump may also affect other secondary ion transporters, including Na
+, Ca
2+ exchange, leading to membrane depolarization[
15]. The PTX-induced membrane depolarization interferes with some vital functions of the cells. Altered concentration of intracellular cations, in particular calcium increase, is generally associated with cell death[
16]. As a consequence of alterations in ion gradients, many modifications of cytosolic proteins occur. Thus, PTX causes modulation of mitogen-activated protein kinase (MAPK) cascades[
6] and stimulates JNK activation in mouse 3T3 fibroblasts[
17]. It was suggested that PTX is also capable of perturbing growth regulatory systems by down-regulation of epidermal growth factor (EGF) binding through a protein kinase C-independent pathway. Inhibition of EGF binding is highly dependent on extracellular Na[
18,
19]. On the other hand, PTX was found to be a non-12-O-Tetradecanoylphorbol 13-acetate (TPA)-type tumor promoter[
20,
21] inducing a signal pathway leading to activation of stress-activated protein kinases (SAPK) JNK important for signal transduction pathways[
22].
The effects shown in various animal species after PTX treatment
in vivo[
23] have been confirmed also by
in vitro studies. Initial experiments were performed on excitable cells of various origin, from muscles and nervous system, and in those cases PTX could be distinguished from other toxins on the basis of severe effects including contractile action on vascular smooth muscle[
24], increase in cation permeability and depolarization[
25] as well as plasma membrane lysis[
16].
The objective of this study was to analyze the effects of PTX on several HNSCC cell lines in comparison to healthy epithelial cells and determine how sensitive xenografted tumors are to this toxin. We also focused on signalling complexes and molecular compounds such as the MAPK/JNK signalling cascades pathway aiming to understand the underlying molecular mechanisms responsible for the difference in PTX toxicity between normal- and HNSCC cells.
Methods
Human cell lines
For the in vitro experiments cell lines derived from human HNSCC of different localizations were used: oropharynx, (UKHN-1), esophagus (UKHN-2), tongue (UKHN-3), and tonsil (UKHN-6). All cell lines were authenticated by single tandem repeat DNA typing (DSMZ, Braunschweig, Gemany). Five human epithelial cell cultures, derived from normal mucosa of the pharynx and the larynx served as controls. Prior to analyses all cell lines were negatively tested for mycoplasma. The normal mucosa biopsies were retrieved during surgery after written patient consent was obtained, in accordance with the Ethical Commission of the University of Schleswig-Holstein, Campus Kiel, subjected to the Helsinki Declaration, revised 1983 (No. AZ D 438/10). The carcinoma cells were grown in minimum essential medium with 10% (v/v) fetal calf serum (Biochrom, Berlin, Germany) at 37°C in 5% CO2 humidified atmosphere. Normal epithelial cells were grown in SFM-medium (Life Technologies, Inc. Eggenstein, Germany) under same conditions as the carcinoma cells.
RT-PCR and Northern blot analysis
Total RNA of tumor and normal cells was extracted using the RNeasy Kit (Qiagen, Hilden, Germany) and 100ng total RNA were used for cDNA synthesis using MMLV reverse transcriptase following the manufacturer’s instruction (Life Technologies, Inc.). For RT-PCR analysis, the cDNA was amplified for 30 cycles using the sense 5’-AGATTCCGAGAAGAAGACCA-3’ and anti-sense 5’-GCTGGGGCTCAGACTCCCCCGTGAGA-3’ oligonucleotides specific for the ATP1AL1 gene (annealing temperature 55°C) and the sense 5’-CCAGCCGAGCCACATCGC-3’ and anti-sense 5’-ATGAGCCCCAGCCTTCTCCAT-3’ oligonucleotides specific for the GAPDH gene (annealing temperature 55°C). For Northern hybridization a 535 bp probe was designed using the sense primer 5‘-CTGTATTGCAGCTAAGCTC-3‘ and the anti-sense primer 5‘-CTACATCTTCAAGGGTCTG-3’ (positions 1799–1817 and 2315–2333) of the JNK3-mRNA (GenBank accession no. NM_002753) which was labelled by chemoluminescence (Roche Diagnostics, Germany). Twenty microgram total RNA from tumor cells and normal epithelial cells were separated on 1% agarose gel. After transfer of the RNA onto nylon membrane both hybridization- and detection procedures were carried out according to the manufacturer’s instructions.
Isolation of PTX
PTX was isolated chromatographically from the marine Cnidaria (zooanthid)
Palythoa caribaeorum and purified as we described earlier[
26]. Purified PTX was lyophilized and stored at −20°C.
Cytotoxicity assay
Quantification of cell death and cell lysis was based on the measurement of LDH activity released from the cytosol of damaged cells into the supernatant using a non-radioactive LDH detection kit (Roche Diagnostics, Germany). Cells grown to a monolayer were incubated for 24 h in the presence (1-4 ng/ml) or absence of PTX. After centrifugation at 250xg for 10 min. the cell-free culture supernatants were collected from PTX treated -and untreated cells and incubated according to the manufacturers instruction. To calculate percent cytotoxicity appropriate controls were measured in each experiment. Absorbance was measured at 492 nm and 620 nm using an ELISA reader (Dynatech MR5000, Denkendorf, Germany).
Clonogenic assay
At day 0, HNSCC cells and normal epithelial cells were plated in duplicate in 6-well plates. One week later, after cells had reached confluency, they were incubated for 24h at various PTX-concentrations (0.5 ng/ml to 10 ng/ml). Subsequently, cells were washed with PBS, fixed in ethanol and stained with crystal violet (0.1% crystal violet in 20% ethanol). Stained cells were measured by microscopic counting randomly choosing at least ten middle power magnification (x200) fields. Mean values and standard deviation (SD) were calculated.
JNK3 inhibitory assay
Pyrazolourea, a selective inhibitor of JNK3 (IC50 = 7 nM[
27]) was obtained from Merck/Calbiochem, Germany. Normal epithelial cells were seeded in 6-well plates and cultured until confluent. The cells were incubated with pyrazolourea at concentrations ranging from 20 nM to 100 nM for 3 hours to inhibit the JNK3 protein kinase. Subsequently, cells were exposed to 6 ng/ml PTX for 24 hours. Finally, cell survival was determined using the crystal violet assay.
Animal experiments
SCID bg/bg mice were obtained from Charles River (Sulzfeld, Geramny) aged 10 to 12 weeks. For the carcinogenicity experiments a group of tumor-free mice (n=8) was treated by subcutaneous (sc) injection of 0.5ng PTX in a volume of 20 μl PBS/day for 5 days. Subsequently,the animals were observed over a period of 8 months. After that period the internal organs such as liver, kidneys, and spleen were examined histopathologically.
For the therapeutic response study mice were divided into three tumor groups and a control group. In the tumor groups (n=24) mice were injected sc in flank with one million tumor cells (a mix of UKHN-1, UKHN-2, and UKHN-3 cell lines 1:1:1). Two weeks after tumor cell inoculation, one tumor group (n=8) received intratumoral injections (1,5 ng PTX dissolved in 20 μl PBS ) every three days over a period of further 24 days, another tumor group (n=8) received intraperitoneal (ip) injections of the same dose every three days and the third tumor group (n=8) received intratumoral injections of PBS (20 μl). Tumor size was measured three times per week with a calliper. After the therapeutic response study residual tumors as well as liver, kidneys, and spleen were examined histopathologically.
In the control group (n=8) mice were each injected sc in their flanks with one million normal epithelial cells. This group was also observed over the period of further 24 days in order to assure that no tumor growth occurred. The experiments in SCID mice were approved by the Ministry of Environment, Nature and Agriculture of Schleswig-Holstein, Germany. However, with a restriction in animal numbers to be used that led us to use the above mentioned combination of HNSCC cells, rather than using each cell line individually.
Immunohistochemistry
For immunohistochemical evaluation 8 subcutanous xenograft tumors were used. One of them was analysed prior to the start of the intratumoral PTX treatment, one after 8 and 16 days respectively, and the remaining five tumors 24 days after PTX treatment. The tumors were fixed in formalin and embedded in paraffin. Deparaffinized sections (5 μm) were stained with hematoxilin and eosin.
Statistical analysis
Statistical analysis of the data was performed by means of One-Way ANOVA (SPSS vs 20). Data were considered statistically significant if p ≤ 0.05.
Discussion
The data presented here show that normal epithelial cells can tolerate higher concentration of PTX without apparent harm than HNSCC cells. The effect of PTX shown on tumor cells suggests that their morphology can be used as an index of PTX toxicity. Morphological change in tumor cells also correlated with LDH release indicating a loss of cellular function, primarily the membrane integrity as would be expected in response to PTX which is known to affect the plasma membrane[
15]. It is obvious that many of the pharmacological effects of PTX are attributable to the effect of this substance on trans-membrane ion transfer. PTX has a unique action on the Na
+,K
+-ATPase, converting the pump into an ion channel and resulting in K(+) efflux, Na(+) influx and membrane depolarization[
25]. PTX can in vitro cause lysis of mouse spleen cells which has been attributed to a PTX-induced increase in cellular calcium levels[
28].
The toxicity of PTX in mammals is strongly dependent upon the route of administration[
29]. PTX is most toxic by intra venous (
iv) injection, the LD50 in mice amounted to 0.15 -0.53 μg/kg[
1,
30]. The PTX toxicity by
ip administration is lower than that by iv injection, with values of 0.31-1.5 μg/kg being reported for mice[
31,
32]. PTX is much less toxic orally than after
iv or
ip administration. Results from the few existing studies reports an oral LD50 from 510 μg/kg to 767 μg/kg in mice[
33,
34].
PTX has been described as a tumor promoter[
20,
21]. This might misleadingly suggest that it is capable of causing tumors. Therefore it is important to note that the basis to classify an agent as a tumor promoter is conditional and is performed only within the context of a two-stage (or multistage) model protocol[
35]. The tumor-promoting activity of PTX has been investigated earlier using mouse skin[
20]. Thereby, in the first stage carcinogenesis was initiated with the mutagenic compound 7,12-dimethylbenz[a]anthracene (DMBA). In the second stage, repeated application of PTX was performed over a period of several weeks. In mice treated with DMBA and PTX tumor development occurred, but no tumors were observed in animals treated with PTX alone suggesting that PTX treatment alone is not sufficient to generate tumors. To verify this, we performed long-time experiments in which a group of mice were treated daily with 0.5ng PTX for 5 days. By using this low PTX concentration we based our approach on results which showed that PTX concentrations higher than 0.5ng are already toxic to mice[
36]. The animals were observed over a period of 8 months without finding evidence of tumor development. Also other studies showed that PTX does not act as a tumor initiator in a Balb/c 3T3 cell transformation assay[
37] and it was negative in the Ames mutagenecity test using different strains (TA98, TA100, TA102 and TA1537)[
29]. Based on these findings we used PTX to treat tumor xenografts established in SCID mice. Treating these mice with doses as little as 68-83ng/ kg bodyweight we observed rapid and progressive tumor destruction without recognizing any apparent disease symptoms. However, this was only the case when PTX was admistisred intratumoral. None of the mice did show any undesired pattern of behavior during therapy nor during a follow up period of 2 weeks, suggesting that low doses of intratumoral injected PTX might even be beneficial, due to them selectively killing tumor cells rather than normal epithelial cells, but no effects were seen after ip PTX injections.
Alterations in ion gradients induced by PTX at the plasma membrane level play a crucial role in cytotoxic and cell death events[
38]. Experimental studies indicated that PTX targets the Na
+, K
+ ATPase, and thereby destroys the ion gradient [5,9,12,]. This may lead to a lack of Na
+, K
+ ATPase causing dramatic effects on cell function. It is reasonable to hypothesize that a response of the cells to this external influence is the post-production of Na
+, K
+ ATPase in order to replace the quantity indispensible for stable cellular conditions. To demonstrate this we analyzed the transcriptional activity of several genes and found that treatment of cells with PTX in fact influences the expression of the ATP1AL1 gene that encodes the Na
+, K
+ ATPase. The initial down-regulation and the subsequent progressive up-regulation of this gene is a typical phenomenon of self-regulating, self-protection processes i.e. the ability of the cells to maintain their internal equilibrium due to PTX as an external influencing factor. PTX on the other hand seems to influence the energy metabolism of the cells since we have shown that GAPDH gene expression was also down –and up-regulated as a function of PTX concentration. The expression profiles for both ATP1AL1 and GAPDH genes suggest that PTX induces in the cell lines studied both transcriptional gene suppression and activation. The mechanism involved in such bidirectional transcription process is poorly defined. Recent observations suggest that bidirectional transcription in human cells is an endogenous gene regulatory mechanism whereby small non-coding RNA mediated transcriptional regulation can act in both suppressive[
39] and activating manner[
40].
PTX stimulates JNK activation through a pathway that involves ion flux[
6]. Initial studies showed that PTX affects JNK activation through a mechanism that involves sodium influx[
17]. A later study conducted in rat fibroblasts suggested that PTX stimulates JNK activation through a mechanism that involves potassium efflux[
41]. It was also demonstrated that PTX-stimulated signals are transmitted to JNK through the activation of a protein kinase cascade, so that the induction of ion flux by PTX results in the activation of MEK4 (MAPK kinase 4) which phosphorylates and activates JNK[
42‐
45]. Collectively, the JNK MAPKs as an evolutionarily-conserved family appear to be important mediators of PTX-stimulated signals. Noteworthy in this regard is the involvement of JNK3 (MAPK10) in these signaling events and has been verified by our JNK3 protein kinase inhibition experiment showing that the repression of the JNK3 expression is essential for the enhancement of PTX toxicity in cancer cells.
In conclusion, we have demonstrated that head and neck cancer cells and xenografts are more sensitive to PTX than normal cells. Because PTX binds to cell surface receptors present on malignant and benign cells, and acts more effectively upon HNSCC cells, there is a need to pay more attention to this natural product to further define the way of its optimal potential use which may extend our knowledge of the biology of head and neck cancer.
Acknowledgements
We thank Dr. D. Emme (Department of Experimental Cancer Research, Division of Molecular Oncology, University of Kiel Schleswig-Holstein, Germany) for expert animal care; Dr. Dr. h.c. W.G. Forssmann (Immunology and Rheumatology, Medical School (MHH), Hannover, Germany) fore scientific support and A.M. Roen for technical assistance. This study was supported in part by grant GZ: GO 1094/2-1 from the German Research Society.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
T.G. performed the experiments and wrote the manuscript. L.B. purified and isolated the palytoxin and analyzed its toxic effect. E.S.Q. performed the experiments with the control mice and discussed results. P.A. supported the experiments and conducted data analysis. M.H. gave constructive support, collected the data and wrote the paper. All authors discussed the results at all stages. All authors read and approved the final manuscript.