This article is mentioned in:

Introduction

Head and neck squamous cell carcinoma (HNSCC), arising in the oral cavity as well as in the pharyngeal and laryngeal regions of the neck, is the sixth most commonly diagnosed cancer worldwide. In general, complete cure of HNSCC is less than 50% (1). Despite advanced treatment strategies, the outcome of head and neck cancer has not significantly improved to date (2). For this reason, there is still a demand for new substances suppressing tumor growth and invasion.

Chelidonium majus, the greater celandine, has been historically used to fight cancer and other diseases. It was described in detail in Jonathan Hartwell's compendium (3) ‘Plants used against Cancer’. The greater celandine gained overwhelming interest when Ukrain, a drug derived from C. majus, was selected for use in cancer treatment during the 1990s. The drug is, according to the manufacturer, a semi-synthetic derivative of the purified alkaloid chelidonine modified with thiophosphoric acid (Thiotepa). Moreover, it has been proposed to kill tumor cells selectively without damaging primary cells and to exhibit an immune modulatory effect (4–6). There are many studies reporting the treatment of different cancers using Ukrain. During the 1990s, a ‘hype’ about Ukrain arose and numerous case reports appeared in the no longer published journal ‘Drugs Under Experimental and Clinical Research’. Studies reported the benefits of Ukrain such as a prolonged life span or even complete remission, without having adverse side effects (7–9). Ukrain was shown to be cytotoxic for a variety of tumor cell lines in vitro (10–13). Furthermore, several groups proved the induction of apoptosis by Ukrain in diverse tumor systems in vitro (10,14).

As it has been shown that chelidonine is not very effective in HNSCC cell lines (15) and there is still a lack of studies of other C. majus alkaloids as possible therapeutic agents in HNSCC, the aqueous extract NSC-631570 was chosen to be tested in vitro for its action in HNSCC.

Combined liquid chromatography and mass spectroscopy of NSC-631570 was applied. The effects of NSC-631570 and its major alkaloid chelerythrine on drug-sensitive and drug-resistant HNSCC cell lines and on primary mucosal keratinocytes and fibroblasts were investigated in the present study. Apoptosis as well as the influence of the extract on gene expression and on the motility of HNSCC cells in a 3-dimensional, spheroid-based invasion assay were analyzed. The results are discussed critically with respect to previously published research.

Materials and methods

Reagents

NSC-631570 was kindly provided by Dr Wassil Novicky (Vienna, Austria), who is the inventor of Ukrain. Reference substances chelidonine, allocryptopine, chelerythrine and sanguinarine were purchased from Sigma-Aldrich; Merck KGaA.

Liquid chromatography-mass spectroscopy of NSC-631570

LC-MS analysis was performed using a Shimadzu LC-MS-2020 mass spectrometer (Shimadzu Deutschland GmbH) containing a DGU-20A3R degassing unit, a LC20AB liquid chromatograph and SPD-20A UV/Vis detector. A Synergi 4 U fusion-RP column (150×4.6 mm; Phenomonex) was used as a stationary phase, and a gradient of MeOH/water was applied as a mobile phase: Solvent A: water with 0.1% formic acid, solvent B: MeOH with 0.1% formic acid. Solvent A ranged from 0 to 100% in 8 min, and then remained at 100% for 5 min, before reducing from 100 to 5% in 1 min, and then being held 5% for 4 min, The method was performed with a flow rate of 1.0 ml/min, and UV detection was measured at 245 nm. Allocryptopine, chelerythrine, sanguinarine and chelidonine were used as reference substances.

Cell lines and cell culture

Squamous cell carcinoma cell lines, originating from laryngeal or hypopharyngeal tumors were used for the study, comprising the major proportion of cases treated at the ENT, Wuerzburg University Hospital. The cell line FaDu (LG standards) originating from a hypopharyngeal carcinoma was grown in RPMI-1640 medium (Seromed), supplemented with 10% foetal bovine serum (FBS). HLaC78 and HLaC79 cell lines derived from larynx carcinoma (16) were maintained in RPMI-1640 medium. HLaC79-Tax was obtained by isolation and the selective cultivation of a paclitaxel-resistant HLaC79 clone. It was grown as the original cell line supplemented with 10 nM (HLaC79-Tax) paclitaxel. As primary cells, human mucosal fibroblasts and keratinocytes were used. Fibroblast and keratinocyte cultures used, were thawed from frozen samples, generated from a tonsil surgery specimen in 2007. Fibroblast cultures were established by explant cultures in DMEM/10% FBS. The isolation of keratinocytes was performed as previously described (17). Keratinocytes were maintained in Keratinocyte Medium 2 (PromoCell). Patient informed written consent was obtained prior to the study. The present study was approved by the Institutional Ethics Committee on Human Research of The Julius Maximilian University of Wuerzburg (study approval no. 16/06).

Cell viability and proliferation assay

Cells were seeded at 5,000 cells/well in 96-well plates. The cells were then treated with increasing concentrations of NSC-631570, chelerythrine or allocryptopine for 48 h. Cell proliferation was measured 48 h later by administration of MTT at a concentration of 1 mg/ml. After a 4-h incubation, MTT-staining solution was replaced by isopropanol and the cells were incubated at 37°C for 45 min. The color change of yellow MTT to a blue formazan dye was measured with an ELISA reader at a wavelength of 570 nm. Relative toxicity was calculated as the surviving cell % by setting solvent-treated control cells to 100%.

Apoptosis

FACS analysis was performed using the BD Pharmingen Annexin V-APC kit (BD Biosciences) according to the kit manual. In brief, HLaC78 and FaDu cells were treated with EC50 concentrations of NSC-631570 or chelerythrine, respectively for 24 h, harvested and washed twice with cold PBS. The shortened incubation time was chosen in order to enrich early apoptotic stages within the cell populations (18). Cells were then resuspended in 1X binding buffer (0.1 M HEPES, pH 7.4, 1.4 M NaCl, 25 mM CaCl2) at a concentration of 1×106 cells/ml. To 100 µl of this cell suspension, 5 µl Annexin V-APC and 5 µl 7-AAD (included in the kit) were added, the cells were vortexed and incubated for 15 min in the dark. An amount of 400 µl of 1X binding buffer was added. Within 1 h, FACS analysis was performed at an excitation wavelength of 650 nm.

In vitro invasion assay

Tumor spheroids were established by dispending 5,000 cells/well of HLaC78 cells on ultra-low-attachment (ULA) 96-well round-bottomed plates (Corning, Inc.). For the migration assay, spheroids of HLaC78 were transferred manually to different extracellular matrix substrates. For this reason, the surface of flat-bottomed 96-well plates was coated with 0.1% gelatine, 5 µg/ml fibronectin, 50 µg/ml laminin, 50 µg/ml collagen I (all purchased at Sigma Aldrich; Merck KGaA) or 125 µg/ml Matrigel® (Becton Dickinson) for 2 h at room temperature. Wells were washed twice with PBS and blocked with 1% bovine serum albumin in PBS for 1 h; 72-h old spheroids were then manually transferred to the coated wells. Spheroids were incubated with or without NSC-631570 or chelerythrine. Migration was analyzed by photographing spheroids after 1 and 24 h with a Leica DMI 4000 inverted fluorescence microscope (Leica Microsystems) at 5-fold magnification. Incubation time was chosen due to the fact, that the cell line HLaC78 needs approximately 24 h to invade completely without inhibiting substances (personal observation). Quantification of migrated areas was performed with ImageJ software (version 1.52a; National Institutes of Health, NIH, Bethesda, MD, USA).

RNA extraction and RNA quality control

RNA of HNSCC cell cultures was isolated with the RNeasy kit (Qiagen) according to the manufacturer's instructions. Purity and concentration were determined photometrically. In expression arrays, RNA quality was determined with the RNA 6000 Nano kit and a Bioanalyzer 2100 instrument (Agilent). RNA integrity numbers (RINs) of our samples ranged between 9.4 and 9.7.

Microarray analysis

For microarray hybridization, 100 ng total RNA was amplified and labelled using the IVT Express kit and hybridized to GeneChip PrimeView Human Gene Expression arrays (both from Affymetrix; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Raw microarray data were background corrected, normalized and summarized to probe set expression values using the Robust Microarray Average (RMA) algorithm (19,20). Data pre-processing and calculation of fold-changes between treated and untreated expression values was performed with the Affymetrix Transcriptome analysis Console 4.0.1 (Affymetrix; Thermo Fisher Scientific, Inc.). Microarray data were deposited in MIAME-compliant form at Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) with the identifier GSE115874.

TaqMan real-time PCR

In order to confirm the strong increase in CYP1A1 expression caused by NSC-631570 in FADU cells and other HNSCC cell lines, FaDu and HLaC78 cell lines were treated for 48 h with either chelerythrine, allocryptopine or NSC-631570 at their EC50 concentrations. RNA was isolated (see above section) and real-time TaqMan PCR (Applied Biosystems; Thermo Fisher Scientific, Inc.) was performed in triplicates on a real-time PCR cycler (Applied Biosystems; Thermo Fisher Scientific, Inc.) using the TaqMan gene expression assay for CYP1A1. Relative quantification was calculated according to the 2−ΔΔCq method (21). Expression values were normalized to the expression of GAPDH as an endogenous control, stably expressed in HNSCC cell lines.

Tube formation assay

Using tube formation assays, the ability of endothelial cells to form three-dimensional capillary-like structures was analyzed. Ibidi angiogenesis slides (15-well, Ibidi GmbH) were coated with growth factor reduced basement membrane extract (BME; Trevigen). BME gels were overlaid with 1×104 human umbilical vein endothelial cells (HUVECs) in growth medium, with and without the addition of NSC-631570 or chelerythrine, respectively. Cells were incubated for 6 h and images were captured. Analysis of the images was performed by Wimasis. For quantification, four parameters were analyzed: tube length, number of branching points, covered area and number of loops.

Statistical analysis

All statistical analyses and graphs were performed with Graph Pad Prism 6 (GraphPad Software, La Jolla, CA, USA). Data are presented as the mean of three independent experiments or 8 measured spheroids ± standard deviation. As statistical tests unpaired t-test (angiogenesis, migration measurements) or ANOVA Dunnett's multiple comparison test (RT-qPCR) were used. Differences were considered to be significant as indicated in the figures and legends: ****P<0.0001, ***P<0.001, **P<0.01, *P<0.05.

Results

LC-MS analysis

LC-MS analysis of the NSC-631570 extract displayed four major peaks at 6.5, 6.6, 7.5 and 7.7 min in the UV chromatogram at 254 nm (Fig. 1A) with the percentage area of 10.9, 13.8, 21.7 and 43.9%, respectively (Fig. 1B). Corresponding molecular masses are shown in Fig. 1C.

Figure 1. (A) UV-chromatogram of NSC-631570. y-axis, mAU (milli-absorbance units); x-axis, retention time in min. (B) Table of eluted peaks of NSC-631570. (C) Major peaks of LC-MS analysis of NSC-631570 with corresponding molecular masses.

LC-MS of the reference substances revealed molecular masses of 369.05 for allocryptopine, 347.01 for chelerythrine, 331.05 for sanguinarine and 352.95 for chelidonine. According to the present analysis, the most abundant constituents of NSC-631570 are chelerythrine (43.86%) and allocryptopine (24.72%). Their effective concentrations in an extract solution of 10 µg/ml were calculated to be 2.47 µg/ml for allocryptopine and 4.39 µg/ml for chelerythrine.

Cytotoxicity

The cell lines FaDu, HLaC78, HLaC79, HLaC79-Tax, primary mucosal keratinocytes (MKs) and fibroblasts were incubated with increasing concentrations of NSC-631570 (10, 25, 50, 100 µg/ml) for 48 h. The MTT assay was used to measure cell viability and cytotoxicity (Fig. 2). For the calculation of inhibition rates, at least three independent experiments were carried out.

Figure 2. Cytotoxic effect of NSC-631570 at increasing concentrations on HNSCC cell lines HLaC78, FaDu, HLaC79, HLaC79-Tax and primary mucosal keratinocytes (MKs) and fibroblasts as assessed with the MTT assay. Data are presented as the mean of three independent experiments ± standard deviation. Left panels, dose-response curves; right panels, statistical curve fitting/calculation of EC50 (response vs. agonist, variable slope).

NSC-631570 treatment showed clear dose-response curves with calculated EC50 concentrations varying around an average concentration of 10 µg/ml (FaDu 10.99, HLaC78 9.34, HLaC79 12.4, HLaC79-Tax 9.62, MKs 10.65, fibroblasts ~10.47 µg/ml).

P-glycoprotein-overexpressing HLaC79-Tax cells were similarly affected by NSC-631570 as their original paclitaxel-sensitive cell line HLaC79. MKs were also sensitive to NSC-631570, while fibroblasts proved to be much more resistant, even at high concentrations (Fig. 2).

Chelerythrine exerted strong cytotoxicity on HNSCC cell lines FaDu and HLaC78 with an EC50 dose of approximately 3 µM (Fig. 3). Allocryptopine, however, revealed only weak cytotoxic effects on HNSCC cell lines, even at non-physiological doses up to 500 µM (Fig. 3).

Figure 3. Cytotoxic effect of chelerythrine on HNSCC cell lines and primary cells, as well as allocryptopine on HNSCC cell lines FaDu and HLaC78 at increasing concentrations, as assessed using the MTT assay. Left panels, dose-response curves; right panels, statistical curve fitting/calculation of EC50 (response vs. agonist, variable slope). Calculated EC50 values are indicated in the corresponding graphs. Data are presented as the mean of three independent experiments ± standard deviation. MKS, primary mucosal keratinocytes; FBs, fibroblasts.

Apoptosis

Apoptosis of NSC-631570 (NSC) or chelerythrine treated (CE) and untreated (Co) FaDu and HLaC78 cells was determined after 24 h incubation by FACS analysis. Cell lines were incubated with the calculated EC50 doses of each drug. Flow cytometry analysis using the APC Annexin V kit showed an apoptosis-promoting effect for NSC-631570 in both cell lines. With the EC50 dose of 10 µg/ml, NSC-631570 revealed a discrete apoptotic cell fraction of 13.1% in FaDu and 13.5% in HLaC78 cells, respectively (Fig. 4).

Figure 4. FACS analysis of early apoptotic events using the APC Annexin V kit for NSC-631570 (NSC) or chelerythrine (CE) treated FaDu and HLaC78 cells. FaDu and HLaC78 cells were incubated for 24 h with the EC50 of NSC-631570 (10 µg/ml) or chelerythrine EC50 (3 µM). Results are summarized in the diagram below. Co, untreated control cells.

Chelerythrin had a stronger effect with regard to triggering apoptosis, especially in HLaC78 cells. The corresponding EC50 concentrations achieved 15.3% (FaDu) and 28.3% (HLaC78) of cells at the pre-apoptotic stages (Fig. 4).

Cell motility on extracellular matrix (ECM) proteins

Spheroid-based experiments were used to investigate the influence of NSC-631570 on invasion on different ECM substrates.

Spheroids of FaDu and HLaC78 cells were grown in ultra-low-attachment-plate wells. After 72 h, they were pipetted into wells, coated with gelatine, fibronectin, laminin, collagen I and Matrigel and incubated with (Uk) or without (Co) NSC-631570 or chelerythrine (CE) for 18 h. Cell migration was quantified by photographing native and outgrown spheroids at t=0 and t=18. Outgrown areas were measured with ImageJ software (area calculation). For each condition, eight spheroids were measured. Calculation of cell motility was achieved by setting the area at t=0 at 100%. Results are displayed in Fig. 5.

Figure 5. Migration of NSC-631570 (Uk) or chelerythrine (CE)-treated and control (Co) HLaC78 cells on different ECM protein-coated surfaces. Spheroids were treated with respective EC50 concentrations for 24 h. Migration of the cells was determined setting spheroid area of t=0 at 100% percentage of migration of NSC-631570/chelerythrine-treated HLaC78 cells. Data are presented as the mean of eight spheroids ± standard deviation. ****P<0.0001, ***P<0.001, *P<0.05 statistically significant values (unpaired t-test) compared to the control.

The comparison of migrated areas revealed a strong suppression of motility in FaDu cells induced by NSC-631570 on all tested ECM surfaces. Invasion of the highly invasive HLaC78 cell line was also significantly suppressed with the exception of laminin (Fig. 5). Chelerythrine showed no anti-invasive influence in the spheroid-based migration assay (Fig. 5).

Gene expression

To generate a gene expression profile of the NSC-631570-treated cells, microarray analysis was performed using FaDu cells which were treated for 48 h with the EC50 of NSC-631570 (10 µg/ml). Of a total number of 49,372 genes tested, 223 (0.45%) were upregulated and 696 (1.41%) were downregulated by NSC-631570. The top 50 upregulated and downregulated genes are summarized in Table IA and B. Complete array results have been deposited at Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) with the identifier GSE115874. Pathway analysis revealed major expression changes affecting apoptosis, cell cycle, integrin-mediated cell adhesion, RNA processing and translation factors, as summarized in Table II.

Table I. The top 50 upregulated and downregulated genes in FaDu cells triggered by treatment with 10 µg/ml NSC-631570 for 48 h.

Table I.

The top 50 upregulated and downregulated genes in FaDu cells triggered by treatment with 10 µg/ml NSC-631570 for 48 h.

A, Upregulated genes
FC	Gene symbol	Description
23.44	CYP1A1	Cytochrome P450, family 1, subfamily A, polypeptide 1
8.00	CYP1B1	Cytochrome P450, family 1, subfamily B, polypeptide 1
5.61	XAF1	XIAP associated factor 1
5.35	MX2	MX dynamin-like GTPase 2
5.01	ISG20	Interferon stimulated exonuclease gene 20 kDa
4.61	OTUB2	OTU deubiquitinase, ubiquitin aldehyde binding 2
4.37	OAS2	2′-5′-Oligoadenylate synthetase 2
4.26	NFATC4	Nuclear factor of activated T-cells, calcineurin-dependent 4
4.12	OAS1	2′-5′-Oligoadenylate synthetase 1
4.07	SPOCK1	Sparc/osteonectin
3.99	IRF9	Interferon regulatory factor 9
3.94	ALDH3A1	Aldehyde dehydrogenase 3 family, member A1
3.80	SERPINB2	Serpin peptidase inhibitor, clade B (ovalbumin), member 2
3.78	IFI44L	Interferon-induced protein 44-like
3.51	RHCG	Rh family, C glycoprotein
3.44	NAV3	neuron navigator 3
3.43	COL4A6	Collagen. type IV, α 6
3.39	FA2H	Fatty acid 2-hydroxylase
3.31	MDK	Midkine (neurite growth-promoting factor 2)
3.29	DHX58	DEXH (Asp-Glu-X-His) box polypeptide 58
3.28	PCDHB2	Protocadherin β 2
3.26	MYEOV	Myeloma overexpressed
3.25	CORO2A	Coronin, actin binding protein, 2A
3.23	GCHFR	GTP cyclohydrolase I feedback regulator
3.23	ZNF436-AS1	ZNF436 antisense RNA 1
3.21	KRT6A	Keratin 6A, type II
3.20	S100P	S100 calcium binding protein P
3.18	VAT1	Vesicle amine transport 1
3.16	RARRES3	Retinoic acid receptor responder (tazarotene induced) 3
3.16	GALNT12	Polypeptide N-acetylgalactosaminyltransferase 12
3.16	LAMP3	Lysosomal-associated membrane protein 3
3.09	IRF7	Interferon regulatory factor 7
3.06	PRR15	Proline rich 15
3.01	SLFN5	Schlafen family member 5
2.97	C1orf116	Chromosome 1 open reading frame 116
2.96	PTPRM	Protein tyrosine phosphatase, receptor type, M
2.96	ABCG1	ATP binding cassette subfamily G member 1
2.95	SQRDL	Sulfide quinone reductase-like (yeast)
2.93	ADAM8	ADAM metallopeptidase domain 8
2.92	SH3KBP1	SH3-domain kinase binding protein 1
2.92	RSAD2	Radical S-adenosyl methionine domain containing 2
2.91	KHDC1L	KH homology domain containing 1-like
2.91	WFDC2	WAP four-disulfide core domain 2
2.90	PLA2G4A	Phospholipase A2, group IVA (cytosolic, calcium-dependent)
2.87	SYNE3	Long intergenic non-protein coding RNA 341
2.85	STC1	Stanniocalcin 1
2.85	GALC	Galactosylceramidase
2.83	NOV	Nephroblastoma overexpressed
2.82	CPEB4	Cytoplasmic polyadenylation element binding protein 4
B, Downregulated genes
FC	Gene symbol	Description
−17.74	ZNF664	Zinc finger protein 664
−11.35	COTL1	Coactosin-like F-actin binding protein 1
−9.95	DDB1	Damage-specific DNA binding protein 1
−9.31	PLK2	Polo-like kinase 2
−8.11	DHFR	Dihydrofolate reductase
−7.32	FKBP9	FK506 binding protein 9
−6.83	CTGF	Connective tissue growth factor
−6.72	DYNLL2	Dynein, light chain, LC8-type 2
−6.43	EIF4H	Eukaryotic translation initiation factor 4H
−6.24	CBX5	Chromobox homolog 5
−6.14	EMP1	Epithelial membrane protein 1
−6.10	FCF1	FCF1 rRNA-processing protein
−6.07	TMPO	Thymopoietin
−6.01	CHP1	Calcineurin-like EF-hand protein 1
−5.99	TOR1AIP2	Torsin A interacting protein 2
−5.94	SAFB	Scaffold attachment factor B
−5.92	ACTN4	Actinin, α 4
−5.80	OS9	Osteosarcoma amplified 9, endoplasmic reticulum lectin
−5.76	SART1	Squamous cell carcinoma antigen recognized by T-cells 1
−5.69	HNRNPA1	Heterogeneous nuclear ribonucleoprotein A1
−5.68	IST1	Increased sodium tolerance 1 homolog (yeast)
−5.52	SF3B3	Splicing factor 3b subunit 3
−5.50	CYB5R3	Cytochrome b5 reductase 3
−5.40	PIM1	Pim-1 proto-oncogene, serine/threonine kinase
−5.36	YY1AP1	YY1 associated protein 1
−5.31	HSPA5	Heat shock 70kDa protein 5 (glucose-regulated protein, 78 kDa)
−5.27	CNN2	Calponin 2
−5.23	FER	fer (fps/fes related) tyrosine kinase
−5.20	MRPS27	Mitochondrial ribosomal protein S27
−5.08	ARFGAP2	ADP-ribosylation factor GTPase activating protein 2
−5.05	LAMP1	Lysosomal-associated membrane protein 1
−5.05	RELA	v-rel avian reticuloendotheliosis viral oncogene homolog A
−5.01	KRT5	Keratin 5, type II
−4.90	QKI	QKI, KH domain containing, RNA binding
−4.84	STIP1	Stress-induced phosphoprotein 1
−4.82	AP1G1	Adaptor-related protein complex 1, γ 1 subunit
−4.74	MCM7	minichromosome maintenance complex component 7
−4.68	AKT3	v-akt murine thymoma viral oncogene homolog 3
−4.63	CTTN	Cortactin
−4.6	FAM98B	Family with sequence similarity 98, member B
−4.57	ZDHHC7	Zinc finger, DHHC-type containing 7
−4.45	EEF2	Eukaryotic translation elongation factor 2
−4.45	HMGA1	High mobility group AT-hook 1
−4.41	COPB2	Coatomer protein complex subunit β 2 (β prime)
−4.4	MYH9	Myosin, heavy chain 9, non-muscle
−4.39	MCM4	Minichromosome maintenance complex component 4
−4.39	POLDIP3	Polymerase (DNA-directed), delta interacting protein 3
−4.37	RAB1A/1B	RAB1A, RAB1B, members RAS oncogene family
−4.32	CNOT1	CCR4-NOT transcription complex subunit 1

[i] FC, fold change.

Table II. Major pathways concerned in FaDu cells, triggered by treatment with 10 µg/ml NSC-631570 for 48 h.

Table II.

Major pathways concerned in FaDu cells, triggered by treatment with 10 µg/ml NSC-631570 for 48 h.

Pathway/FC	Gene symbol	Description
Apoptosis
3.09	IRF7	Interferon regulatory factor 7
2.53	IRF5	Interferon regulatory factor 5
−2.10	IKBKB	Inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase β
−5.05	RELA	v-rel avian reticuloendotheliosis viral oncogene homolog A
−2.18	BAK1	BCL2-antagonist/killer 1
−3.69	CASP2	Caspase 2
−2.78	CASP8	Caspase 8, apoptosis-related cysteine peptidase
−2.34	BCL2L1	BCL2-like 1
−2.56	TP53	Tumor protein p53
−3.49	MCL1	Myeloid cell leukemia 1
−2.47	BCL2L2	BCL2-like 2
Cell cycle
−2.17	CDH1	Cadherin 1, type 1
−2.18	CDC14B	Cell division cycle 14B
2.35	TBC1D8	TBC1 domain family, member 8 (with GRAM domain)
−2.53	HDAC1	Histone deacetylase 1
−2.09	CDC20	Cell division cycle 20
−2.67	PRKDC	Protein kinase, DNA-activated, catalytic polypeptide
−2.09	HDAC8	Histone deacetylase 8
−3.47	PLK1	Polo-like kinase 1
−3.36	SKP2	S-phase kinase-associated protein 2, E3 ubiquitin protein ligase
−4.39	MCM4	Minichromosome maintenance complex component 4
−2.62	MCM6	Minichromosome maintenance complex component 6
−4.74	MCM7	Minichromosome maintenance complex component 7
−2.89	E2F4	E2F transcription factor 4, p107/p130-binding
2.63	CCNE1	Cyclin E1
−3.49	CCND2	Cyclin D2
Integrin-mediated cell adhesion
−2.28	PDPK1	3-phosphoinositide dependent protein kinase 1
−3.09	CRK	v-crk avian sarcoma virus CT10 oncogene homolog
−2.02	SHC1	SHC (Src homology 2 domain containing) transforming protein 1
2.17	SEPP1	selenoprotein P, plasma, 1
2.29	ITGA2	Integrin, α 2 (CD49B, α 2 subunit of VLA-2 receptor)
−2.77	CAPN1	Calpain 1, (mu/I) large subunit
−3.52	PXN	Paxillin
−2.73	PDPK1	3-Phosphoinositide dependent protein kinase 1
−4.68	AKT3	v-akt murine thymoma viral oncogene homolog 3
mRNA processing
−2.05	RPS28	Ribosomal protein S28
−2.12	DHX16	DEAH (Asp-Glu-Ala-His) box polypeptide 16
−2.83	CPSF4	Cleavage and polyadenylation specific factor 4
−3.77	RNPS1	RNA binding protein S1, serine-rich domain
−3.13	SRRM1	Serine/arginine repetitive matrix 1
−2.02	SFSWAP	Splicing factor, suppressor of white-apricot family
−3.14	SUGP1	SURP and G-patch domain containing 1
−2.40	RNPS1	RNA binding protein S1, serine-rich domain
−2.07	NCBP2	Nuclear cap binding protein subunit 2
−3.85	NXF1	Nuclear RNA export factor 1
−2.72	HNRNPL	Heterogeneous nuclear ribonucleoprotein L
−2.88	HNRNPU	Heterogeneous nuclear ribonucleoprotein U
−2.45	TAF13	TAF13 RNA polymerase II
−2.82	POLR1A	Polymerase (RNA) I polypeptide A
−2.79	GTF2E1	General transcription factor IIE subunit 1
−2.71	POLR2E	Polymerase (RNA) II (DNA directed) polypeptide E, 25 kDa
−2.76	ERCC2	Excision repair cross-complementation group 2
Translation factors
−4.45	EEF2	Eukaryotic translation elongation factor 2
−3.58	EIF2S3	Eukaryotic translation initiation factor 2, subunit 3 γ, 52kDa
−2.65	EIF3F	Eukaryotic translation initiation factor 3, subunit F
−6.43	EIF4H	Eukaryotic translation initiation factor 4H
−4.44	EEF2	Eukaryotic translation elongation factor 2
−2.75	EIF4B	Eukaryotic translation initiation factor 4B
−2.57	EIF5A/5AL1	Eukaryotic translation initiation factor 5A; 5A-like 1
−2.24	EIF2AK1	Eukaryotic translation initiation factor 2-α kinase 1
−2.10	EIF4A2	Eukaryotic translation initiation factor 4A2

[i] Genes were clustered using the Affymetrix tool TAC4.0. FC, fold change.

Microarray data validation by TaqMan RT-PCR

Real-time PCR was performed for NSC-631570-treated cell lines FaDu and HLaC78 to verify array results and to analyze whether differential gene expression is a cell-type independent process. For RT-qPCR, TaqMan probes for CYP1A1 and Cyp1B1 were chosen, as these genes were the most upregulated genes and with respect to a previously published study, downregulation of these genes was observed following treatment with a Chelidonium majus extract (22). FaDu and HLaC78 cells were treated with (NSC) or without (Co) NSC-631570 or EC50 concentrations of chelerythrine (CE) or allocryptopine (AC) respectively, for 48 h.

CYP1A1 and CYP1B1 expression was upregulated in both cell lines in response to treatment with NSC-631570, verifying array data and indicating a specific effect of the Chelidonium majus extract, which was not restricted to one cell line (Fig. 6). Chelerythrine and allocryptopine at least contributed to CYP1A1 upregulation in both cell lines as compared to the control (Co). For CYP1B1, upregulation was confirmed in both cell lines following treatment with NSC-631570. Treatment with chelerythrine or allocryptopine showed no significant difference in expression changes compared to the untreated control.

Figure 6. TaqMan RT-PCR analysis of NSC-631570 (NSC), chelerythrine (CE) and allocryptopine (AC)-treated FaDu and HLaC78 cells. Gene expression was measured followed 48 h incubation of the cell lines with EC50 concentrations of each substance/extract. Values are presented as mean ± standard deviation. ****P<0.0001, ***P<0.001, **P<0.01; statistically significant values; ANOVA Dunnett's multiple comparison test, compared to the control. Co, untreated control cells.

Angiogenesis

NSC-631570 was tested for anti-angiogenic properties with the tube formation assay. Tube formation of endothelial cells, naturally occurring on reconstituted basement membrane, was obviously inhibited by the Chelidonium majus extract NSC-631570 (Fig. 7, upper panels). The quantification of inhibition showed significant decreases in all parameters tested (Fig. 7). There was no significant inhibition of any tube formation parameter by chelerythrine (CE) (Fig. 7, lower panels).

Figure 7. Anti-angiogenic action of NSC-631570 (NSC) and chelerythrine (CE) was tested using a tube formation assay. HUVECs were incubated for 6 h with or without NSC-631570 or chelerythrine at EC50 concentrations. Image analysis showed a significant decrease in total tube length, total branching points, total loops and covered area (%) by NSC-631570 treatment compared to untreated HUVEC tube formation control (Co). In contrast, chelerythrine was not able to decrease any of the parameters significantly. Data are displayed as the mean of five tube formation assays ± standard deviation ****P<0.0001, **P<0.01, statistically significant values (unpaired t-test) compared to the control. HUVECs, human umbilical vein endothelial cells.

Discussion

In the present study, the Chelidonium majus extract NSC-631570 was examined for its effect on HNSCC cell lines in vitro. Particularly, the influence of the drug with respect to cell growth, apoptosis, migration and angiogenesis on tumor cells as well as its cytotoxicity against primary cells was investigated.

LC-MS analysis revealed chelerythrine and allocryptopine to be the most abundant alkaloids of the Chelidonium majus extract. Composition of the extract was found to be essentially comparable to the LC/MS analysis data of Ukrain published by Jesionek et al in 2016 (23).

NSC-631570 was demonstrated to exhibit strong cytotoxic activity on HNSCC tumor cells, originating from laryngeal or hypopharyngeal subsites, even on those overexpressing p-glycoprotein. Results may differ for head and neck cell lines, derived from other locations. Primary cells were also affected by NSC-631570; however, sensitivity differed between generally sensitive epithelial keratinocytes and mucosal fibroblasts. Fibroblasts were much more resistant against NSC-631570 compared to the HNSCC cell lines tested in this study. Therefore, this study cannot support the advertised selective toxicity of Ukrain on cancer cells without restriction. In 2013, a study was published, which analyzed the uptake and intracellular effects of celandine alkaloids in normal and malignant cells (24). Obviously, the toxicity of alkaloid mixtures depends on the differential uptake into the different cell types as well as various targets within the cells (e.g. post-translational modified tubulin isotypes), also differing between cell types. Therefore, the exact effect of an alkaloid mixture in differential cell systems seems to be hardly predictable. Chelerythrine showed a comparable toxicity profile on tumor cell lines, but also a harsher effect on primary cells, also on fibroblasts. This is in agreement with the study of Malíková et al (25), who also observed increased sensitivity of primary cells against chelerythrine. Chelerythrine is a known protein kinase C (PKC) inhibitor and in general PKC inhibitors such as staurosporin and others have failed to be effective in cancer therapy, although anticancer activity was exhibited in several in vitro systems (reviewed in ref. 26). Recently, however, new mechanisms of action beyond PKC inhibition have been published for chelerythrine, concerning the influence on gene expression in cancer cells (27,28). These results, together with the observation, that chelerythrine is able to delay the growth of xenografted tumors in vivo without considerable toxic side effects (29), shed new light on the potential use of chelerythrine as an anticancer therapeutic agent.

HNSCC cell death induced by allocryptopine required concentrations up to 1 mM, excluding this alkaloid from further studies. Certain changes in gene expression, however (e.g. upregulation of CYP1A1), triggered by NSC-631570, may be caused or at least supported by the action of allocryptopine. The upregulation of cytochrome P4501A in response to allocryptopine and protopine has been shown in human hepatocytes and hepatic cancer cells (29).

Cell migration, one of the hallmarks of metastasis, is crucial for tumor formation and progression towards metastatic phenotypes. Especially in HNSCC, this process can be life-threatening, as surgical removal of locally invaded tumor cells in surrounding tissue is difficult without severe functional impairment. To evaluate the influence of NSC-631570 on the invasion on gelatine, laminin, fibronectin, collagen type I and Matrigel, a spheroid-based invasion assay was performed. FaDu and HLaC78 spheroids were treated with or without the EC50 of NSC-631570 for 24 h. Invasion of FaDu cells was significantly inhibited on all tested ECM substrates. For the highly invasive cell line HLaC78, significance was not reached for the invasion inhibition on laminin.

Chelerythrine failed to inhibit invasion in our invasion model, although it has been published to have anti-metastatic properties in hepatocellular carcinoma cell lines (30).

Tube formation assays using HUVECs, cultivated with or without NSC-631570, revealed a significant inhibition of angiogenesis, which was previously presented at a congress in Brazil in 1998 (31) by Wassil Nowicky, the manufacturer of NSC-631570. The role of chelerythrine in angiogenesis inhibition has not been analyzed to date, although it has been shown that it was able to downregulate the expression of vascular endothelial growth factor A (VEGF-A) in MCF-7 breast cancer cells. In the present study, chelerythrine clearly failed to inhibit angiogenesis. However, it has to be taken into consideration, that the tube formation assay solely comprises inhibition of endothelial cell tube formation. It does not comprise tumor cell-derived angiogenic signals.

The high impact of NSC-631570 was found to be accompanied by gene expression changes. Microarray analysis showed differential expression of 1,195 genes caused by NSC-631570, whereby 268 genes were upregulated and 927 were downregulated. The most striking changes in gene expression revealed by microarray analysis occurred in the genes CYP1A1 and CYP1B1, which were strongly upregulated. CYP1A1 and CYP1B1 are involved in the xenobiotic metabolism of poly-aromatic carcinogens and steroid hormones (among others), as well as anticancer drugs. Therefore, they have been associated with drug resistance in cancers (32). Microarray data captured for CYP1A1 and CYP1B1 were verified by RT-PCR for FaDu and HLaC78 cells, treated either with NSC-631570, chelerythrine or allocryptopine alone.

The upregulation of CYP1A1 and CYP1B1 occurred in both cell lines, indicating that this effect is not specific for the FaDu cell line and may be a general effect in HNSCC cells. Chelerythrine and allocryptopine seem to contribute to the upregulation of CYP1A1, while CYP1B1 expression was not influenced significantly by both active ingredients. These results are not in agreement with previous studies of El-Readi et al (22). They observed a downregulation of CYP1A1 and CYP1B1 following treatment with a C. majus extract, concluding that the tested colon cancer cells overcame drug resistance upon treatment. The role of the mainly extrahepatic occurring CYP1 family, which belongs to the cytochrome P450 system, and especially that of CYP1A1, is controversially discussed. CYP1A1 has been shown to be involved in carcinogenesis by activating pro-carcinogens, such as N-nitrosamines. On the other hand it has been shown to be cancer-preventive by metabolizing and thus activating dietary compounds, such as flavonoids (reviewed in ref 33). Possibly, a certain homeostasis of the diverse functions of that enzyme determines its role in cancer. The observed upregulation of CYP1A1 in response to treatment with NSC-631570 or the isolated celandine alkaloids chelerythrine and allocryptopine in HNSCC cell lines therefore cannot be assigned to the cytotoxicity of the drug or to a progression in drug resistance. Interestingly, similar results were published by Orland et al (34). The authors likewise reported an upregulation of CYP1 family genes in HepG2 liver cancer cell lines, caused by a Chelidonium majus extract.

In summary, NSC-631570 showed a high anticancer impact in HNSCC cell lines. The extract exhibited cytotoxic, anti-invasive and pro-apoptotic activities on HNSCC cell lines. Furthermore, it acted in an anti-angiogenic manner on endothelial cells. Compared to its major ingredient chelerythrine, it showed clear advantages concerning the aqueous formulation, the toxicity on primary cells and anti-migratory properties.

Acknowledgements

We would like to thank Dr Wassil Nowicky for providing NSC-631570.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

RH performed all cell culture experiments, apoptosis and RT-qPCR. JS performed LC-MS analysis. JR assisted in molecular biological experiments. CP was responsible for cell culture maintenance and tube formation assays. UH provided material and equipment for LC-MS analysis. MS collected the data and evaluated the results and assisted in writing the manuscript. All authors read and approved the manuscript and agree to be accountable for all aspects of the research in ensuring that the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Ethics approval and consent to participate

Mucosal keratinocytes and fibroblasts were obtained from tonsil surgery. Informed written consent was obtained prior to enrolment in the study. The study was approved by the Institutional Ethics Committee on Human Research of the Julius Maximilian University Würzburg (study approval no. 16/06).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests. There is no other active or previous relation or collaboration with Dr W. Nowicky, who generously provided NSC-631570.

Glossary

Abbreviations

Abbreviations:

References