The head and neck squamous carcinoma cell line HLaC79 was established from a lymph node metastase of a laryngeal squamous cell carcinoma (21). The cell line was grown with RPMI-1640 medium (Seromed, Munich, Germany), supplemented with 10% fetal calf serum (FCS). HLaC79 cells were treated with 10 nM paclitaxel. A resistant clone was isolated by selective trypsination of single clones. The permanent HLaC79 clonal cell line HLaC79-Tax was cultured in RPMI-1640 medium, supplemented with 10% FCS and 10 nM paclitaxel. Hep2 cells were grown in MEM medium, supplemented with 10% FCS. A paclitaxel-resistant clone was isolated according to the procedure in HLaC79-Tax.

Human umbilical vein endothelial cells (HUVEC) were isolated as described elsewhere (22). They were grown in ECGM (Provitro, Berlin, Germany) medium.

Dried and cut Galium verum L. leaves (Herba gallii lutei) were kindly provided by Dr Ivo Pischel, PhytoLab GmbH & Co. KG. (Vestenbergsgreuth, Germany). Tea was prepared as follows: 100 ml boiling water were poured over 15 g of dried and powdered Galium leaves. After cooling down, the supernatant was cleared by centrifugation and sterile filtration. Aliquots were frozen at −80°C. One batch of frozen Galium extract was used for all experiments.

Samples were analyzed using thin layer chromatography, a physical method to separate complex chemical substance mixtures on a stationary phase with a mixture of solvents. The method relies on dynamic differential migration of substances in the mixture as a result of adsorption characteristics, molecular weight, and solubility (23). As samples for TLC Galium decoction, a methanolic Galium verum extract (24) and standards were used. They were applied to Silica gel 60 F264 plates (Merckmillipore.com). As solvents ethyl acetate - formic acid - acetic acid - water (100:11:11:27) for flavones or toluene - ethyl acetate - formic acid (50:40:10) for polyphenolcarboxylic acids were used. Compounds were determined and documented by exposure to UV light at 365 nm with a Camag TLC documentation system (Camag.com). The following standards were used: apigenol, luteolin, apigenol-7-O-glucoside, luteolin7-O-glucoside, quercetol and rutoside for flavones; chlorogenic acid, caffeic acid, rosmarinic acid, ferulic acid, and p-coumaric acid for polyphenolcarboxylic acids.

Cells were seeded at 5000 cells/well in 96-well plates. They were treated with increasing concentrations of Galium verum aqueous extract (50 and 100 μl/ml) for 48 h. Controls were kept in medium supplemented with 100 μl/ml water. Cell proliferation was measured after 48 h by replacing the culture medium with medium containing 1 mg/ml MTT. After 4 h of incubation, MTT-staining solution was replaced by isopropanol and cells were incubated at 37°C for 45 min. The colour conversion of MTT to a blue formazan dye was measured with an ELISA reader at a wavelength of 570 nm. The amount of formazan dye is in direct proportion to the number of metabolically active cells in the culture. Relative toxicity was calculated as % surviving cells by setting control cells treated with vehicle as 100% surviving cells.

For western blot analysis, cells were harvested by scraping and dissolved in RIPA (PBS, containing 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS), supplemented with 10 μg/ml phenylmethanesulfonyl fluoride (PMSF). Protein content was determined according to the method of Lowry (25). Equal amounts of total protein lysates were loaded on 10% SDS-polyacrylamide gels and run at a constant current of 20 mA. Gels were blotted onto nitrocellulose membranes according to the semidry method of Kyhse-Andersen (26). Blots were blocked for 1 h with TBST (10 mM Tris, 150 mM NaCl, 0.05% Tween-20, pH 8.0), containing 5% non-fat dry milk. For detection of p-glycoprotein the monoclonal anti-pGp antibody Clone F4 (Sigma-Adrich, Steinheim, Germany) was used. Primary antibodies were incubated overnight at 4°C, after washing 3 times with TBST, cells were incubated with corresponding secondary antibodies, coupled to horseradish peroxidase for 1 h. After washing once again, detection of bound antibody conjugates was performed with the enhanced chemiluminescence system (ECL, Amersham Biosciences, Freiburg, Germany), according to the manufacturer’s protocol.

3D invasion in collagen I/Matrigel^® gels. Tumour spheroids were generated by seeding 5000 cells/well of Hep2 and Hep2-Tax cells on ultra-low-attachment (ULA) 96-well round-bottomed plates (Corning, Amsterdam, The Netherlands) (27). HLaC79 and HLaC79-Tax cells did not form stable spheroids. After 3–4 days medium was replaced by either Matrigel (BD Biosciences, Heidelberg, Germany) or Vitrogen-100 collagen (Cohesion Technologies, Palo Alto, CA, USA) gels, prepared with or without Galium extract (33.3 μl/ml). Matrigel was diluted 1:2 with cell culture medium. Vitrogen-100 gels were supplemented with 10% 10X MEM medium (Life Technologies, Carlsbad, CA, USA) and 5% Sodium Bicarbonate (Life Technologies, 7.5% solution for cell culture). Both gels were further supplemented with 25 ng/ml EGF (Sigma-Aldrich, Taufkirchen, Germany). After solidifying gels at 37°C complete medium with or without Galium decoction (33.3 μl/ml) was added on top. Photographs were taken after 1, 24 and 48 h after embedding with a Leica DMI 4000 inverted fluorescence microscope (Leica Microsystems, Wetzlar, Germany).

For this tumour spheroid-based migration assay, spheroids of Hep2 and paclitaxel-resistant Hep2-Tax spheroids were placed on different extracellular matrix substrates (28). The surface of flat-bottomed 96-well plates were coated with 0.1% gelatin, 5 μg/ml fibronectin, 50 μg/ml laminin, 50 μg/ml collagen I (all from Sigma-Aldrich, Taufkirchen, Germany) or 125 μg/ml Matrigel (Becton Dickinson, Heidelberg, Germany) for 2 h at room temperature. Wells were washed twice with PBS and subsequently blocked with 1% bovine serum albumin in PBS for 1 h. For 2 days on ULA plates pre-cultivated spheroids (see above) of Hep2 and Hep2-Tax cell lines were transferred to the coated wells with a multichannel pipette. Spheroids were incubated with or without Galium decoction (33.3 μl/ml). Migration was recorded by photographing spheroids after 1 and 24 h with a Leica DMI 4000 inverted fluorescence microscope (Leica Microsystems). Quantification of migrated cells was carried out with ImageJ software (National Institutes of Health, NIH, USA).

Cell lines were treated with 33.3 μl/ml Galium verum extract for 48 h. After 48 h, cells were seeded after incubation with Galium extract at equal cell numbers in multiwall plates. Complete medium (MEM or RPMI) was replaced by Optimem (Invitrogen, Karlsruhe, Germany), which is a complete, serum-free medium. Conditioned medium was collected after 18 h. Conditioned (20–60 μl) medium was directly subjected to electrophoresis in 10% SDS-polyacrylamide gels under non-reducing conditions (29), containing 1 mg/ml gelatin (Sigma-Aldrich, Traunstein, Germany). After electrophoresis, gels were renaturated two times for 30 min in 2.5% Triton X-100 and developed overnight in developing solution (50 mM Tris.HCl, pH 6.8, 0.2 M NaCl, 10 mM CaCl₂, 0.02% Brij 35) at 37°C. Subsequently they were stained with Coomassie Brilliant Blue, destained and dried.

Using tube formation assays, the ability of endothelial cells to form three-dimensional capillary-like structures was analyzed. Angiogenesis-slides (15-well, Ibidi GmbH, Munich, Germany) were coated with growth factor reduced basement membrane extract (BME; Trevigen, MD, USA). After polymerization of BME, the gels were overlaid with growth medium containing 1×10⁴ HUVECs and the 33.3 μl/ml Galium extract. Cells were incubated for 6–8 h and images were taken. Evaluation of pictures was performed by Wimasis GmbH (Munich, Germany). For quantification, four parameters were analyzed: tube length, number of branching points, covered area and number of loops.

RNA extraction and RNA quality control. RNA of Galium-treated and untreated cell lines Hep2 and HLaC79 (48 h, 33.3 μl/ml) was isolated with the RNeasy kit (Qiagen, Hilden, Germany) according to the manufacturers instructions. RNA quality was assessed with the RNA 6000 Nano kit using the Bioanalyzer 2100 instrument (Agilent, Böblingen, Germany). RNA integrity numbers (RINs) of our samples ranged between 9.4 and 9.9.

For microarray hybridization, 100 ng total RNA were amplified and labelled using the IVT Express kit and hybridized to GeneChip PrimeView Human Gene Expression arrays (both from Affymetrix, Santa Clara, CA, USA). Microarray washing and staining was performed with a Fluidics Station 450; fluorescence intensities were scanned with a GeneChip Scanner 3000 7G (both from Affymetrix). Image processing (microarray grid definition, feature intensity readout) and CEL file generation was achieved with the GeneChip Operating Software (Affymetrix).

Raw microarray data were background corrected, normalized and summarized to probe-set expression values using the Robust Microarray Average (RMA) algorithm (30,31). For each cell line differentially expressed probe-sets resulting from treatment were detected in each cell line by estimating the false discovery rate (FDR) with an empirical Bayesian methodology employing a log-normal normal data modeling (32). Significant (FDR<0.05) probe-sets were subjected to gene set enrichment analysis using DAVID (33). These analyses were performed in the R environment (http://www.r-project.org) using Bioconductor (http://www.bioconductor.org) packages ‘affy’ and ‘EBarrays’.

To verify results of arrays, quantitative real-time Taqman^R PCR (AppliedBiosystems.com) was performed. Different cell lines (Hep2, Hep2-Tax, HLaC79, HLaC79-Tax, Fadu) were treated for 48 h with Galium verum decoction (33.3 μl/ml) or vehicle. Subsequently cells were rinsed with phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄•2 H₂O, 2 mM KH₂PO₄, pH 7.4) and scraped off for RNA isolation, which was performed with the RNeasy kit (Qiagen) according to the manufacturer’s instructions. The High Capacity RNA-to-cDNA Master Mix (Applied Biosystems, Darmstadt, Germany) was used for cDNA reverse transcription. Real-time PCR was performed in triplicates on a Real-time PCR cycler (Applied Biosystems) using Taqman Gene expression assays for MDR-1, TXNIP, MDM2, JUN, MMP-10 and RND3. Relative quantification was calculated according to the 2^−ΔΔCT method (34). Expression values were normalized to the expression of GAPDH as an endogenous control which proved to be expressed most stably throughout the cell lines.

All statistical analyses and graphs were performed with Graph Pad Prism 4 (Graphpad software, La Jolla, CA, USA)

Qualitative phytochemical analysis for the Galium verum decoction in comparison to a known Galium verum methanolic extract (24) was performed by thin layer chromatography, in order to determine the biologically active compounds (polyphenolcharboxylic acids and flavons).

Chromatographic images with polyphenolcarboxylic acid standards clearly demonstrate the presence of chlorogenic acid in both extracts, more pronounced in the aqueous extract (Fig. 1). Luteolin-7-O-glucoside and rutoside (Fig. 2) are present in pronounced amounts in the flavonoid fraction of the aqueous Galium extract.

Expression of P-Gp in Hep2, Hep-2-Tax, HLaC79 and HLaC79-Tax was tested by western blot analysis of whole cell lysates and by Taqman qRT-PCR. Both methods confirmed overexpression of P-Gp in HLaC79-Tax and Hep2-Tax cell lines. Lowest amount of the mdr-1 transcript was detected in HLaC79 cells (Fig. 2). Adequate protein bands in western blots were clearly visible in the paclitaxel-resistant clones Hep-2-Tax and HLaC79-Tax. Weak signals came up in HLaC79 and Hep2 protein lysates (Fig. 2).

The four cell lines were treated with increasing concentrations of Galium aqueous extract. Cell viability and cytotoxicity of the used drugs were measured with the MTT assay. Mean percentage inhibition was calculated from at least three independent experiments. Galium extract significantly suppressed the growth of all cell lines (Kruskal-Wallis-test, p<0.05). The multidrug resistant clones of HLaC79 and Hep2, however, tolerated high Galium concentrations better than the parental drug-sensitive cell lines (Fig. 3).

3D invasion in semisolid gels. Investigation of invasion and motility experiments was carried out using spheroid-based experiments. First these experiments better reflect the solid tumour-microenvironment interaction. Second the widely used Boyden Chamber assay proved to be not reproducible in the actual system.

Spheroids of all for cell lines were grown in ultra-low-attachment-plate (ULA-plate) wells. While Hep2 and Hep2-Tax cells formed round-shaped compact spheroids, HLaC79 cells produced only loose aggregates, resolving on pipetting. In contrast to all other cell lines HLaC79-Tax cells did not even form loose aggregates. Instead, many tiny spheroids appeared within the ULA wells, with cells only loosely attached to each other (Fig. 4). For this reason the examinations were limited to Hep2 and Hep2-Tax cells for detailed motility and invasion experiments. Galium decoction was used at a concentration of 33.3 μl/ml which proved to be not lethal to the cell lines.

When embedding Hep2/Hep2-Tax spheroids into solidified Matrigel, with or without EGF, the cells tended not to invade gels. In collagen gels, however, cells did invade. Hep2-Tax cell migration was stronger than the chemosensitive original Hep2 cells (Fig. 5). This mode of invasion was inhibited by Galium extract within collagen gels.

For quantification of invasion on different ECM substrates Hep2 and Hep2-Tax spheroids were transferred manually to wells, coated with gelatine, fibronectin, laminin, collagen I or matrigel. Images of the cells were taken after attachment to ECM (t=0) and after 24 h (t=24). For quantification of cells migrating out of spheroids the areas of spheroids at t=0 were substracted from areas measured after 24 h, using ImageJ software (Fig. 6).

For evaluation of cell motility the area at t=0 was set at 100%. The percentage of the migrated area was calculated as the following formula: %   migrated   area = 100 × Δ   Area Area t = 0 whereas   Δ   Area   = Area t = 24   − Area t = 0Migration of Hep2 and Hep2-Tax cells on gelatin, fibronectin, laminin, collagen I and Matrigel substrate with or without Galium was measured. Representative examples for Hep2 and Hep2-Tax cells migrating on collagen I and Matrigel with or without Galium are shown in Fig. 7. Invasion patterns on gelatin, fibronectin and laminin showed a similar optical appearance, although percentage invasion differed notably between the substrates (Fig. 8). Highest motility was observed on Matrigel substrate in chemosensitive as well as in chemoresistant cells.

Comparing percentage of the migrated areas of Hep2 and Hep2-Tax cell lines revealed higher motility in chemoresistant Hep2-Tax cells, albeit statistically significant only for laminin and collagen I (unpaired t-test, p<0.05, Fig. 8) with fibronectin as an exception, where both cell lines migrated approximately equal. Additionally invaded areas between vehicle- and Galium-treated spheroids were compared. Results are shown in Fig. 9.

In both cell lines ECM substrate invasion was inhibited significantly by Galium verum decoction at sublethal doses of 33.3 μl/ml (unpaired t-test, p<0.05). Comparison of the percentage reduction of migrated area caused by Galium in the two cell lines revealed a stronger inhibition of invasion in the aggressively invading chemoresistant Hep2-Tax cells (Table I).

Gelatinolytic activity of Galium-treated cell lines was examined using gelatin zymography. Culture supernatants of Hep2, Hep2-Tax, HLaC79 and HLaC79-Tax cells, treated with sublethal doses of Galium or vehicle were separated and developed on gelatin zymographic gels. Results are shown in Fig. 10. Gelatinolytic activity of the active 66 kDa MMP-2 was diminished in the cell lines HLaC79, Hep2 and HLaC79-Tax by Galium decoction treatment. In the paclitaxel-resistant clone Hep2-Tax no effect of Galium treatment on gelatinolytic activity was detectable.

Effects on angiogenesis were tested with the tube formation assay, quantifying total tube length, total branching points, total loops and % covered area. There was no statistical significant or visible influence of Galium extracts at 33.6 or 66.6 μl/ml on tube formation of HUVECs growing on Matrigel (Fig. 11, here shown for 66.6 μl/ml).

Using the PrimeView Human Gene Expression array (Affymetrix), the expression profiles of two cell lines (HLaC79 and Hep2) following a Galium treatment (33.3 μl/ml, 48 h) in comparison to untreated cells were created. Of 48,736 analyzed genes, 367 were significantly differentially expressed in Hep2 cells and 279 in HLaC79 cells following Galium treatment. There were only 11 genes affected significantly in both cell lines (Table III) in this experiment, indicating that the effect of this cocktail of active substances is dependent on cell type and/or cell status (cell cycle, passage number, gene expression profile, differentiation and/or hormonal stage).

The most differentially expressed genes for both cell lines are listed in Table II. For both cell lines, significantly up/downregulated genes are listed in Table III. AKR1C1 and C3 genes, both coding for proteins needed in detoxification as well as HES1 are oppositionally regulated in Hep2 and HLaC79 cell lines (AKR1C1 HLaC79 logFC= −1.51; Hep2 logFC=3.03; AKR1C3 HLaC79 logFC= −1.43; Hep2 logFC=3.28; HES1 HLaC79 logFC= −2.04; Hep2 logFC=1.75).

To test array results on specificity for Galium treatment a few interesting candidate genes were picked out and qRT-PCR with Galium-treated and untreated cells of the cell lines Hep2, Hep2-Tax, HLaC79, HLaC79-Tax and Fadu was performed. Expression induction/repression of TXNIP, MDM2 and RND3 differentially induced/repressed by Galium extract in Hep2 cells and JUN and MMP-10 in HLaC79 cells was tested. Results are summarized in Fig. 12.

Matrix metalloproteinase 10 (MMP-10) and c-Jun (JUN) were downregulated significantly in HLaC79 cells after Galium treatment. HLaC79 cells expressed vast amounts of c-jun at the stage of RNA isolation for the cDNA array. In the other cell lines, however, qRT-PCR revealed no clear tendency towards up or downregulation of JUN by Galium decoction. It is remarkable, however, that basic expression levels in all other cell lines are much lower than in HLaC79.

MMP-10 was significantly downregulated in HLaC79 cells after Galium treatment. There was no amplification of MMP-10 in Hep2 and Hep2-Tax cells. Expression was diminished in HLaC79 cell line. In Fadu and HLaC79-Tax cells, however, MMP-10 expression increased after Galium incubation (in Fadu significantly, in HLaC79-Tax cells not significantly, unpaired t-test).

Of the genes differentially affected in expression by Galium in Hep2 cells the tumour suppressor gene Thioredoxin-interacting protein (TXNIP) was chosen. TXNIP was significantly upregulated by Galium aqueous extract in Fadu, Hep2 and Hep2-Tax cells. In HLaC79 and HLaC79-Tax cell lines, expressing low amounts (HLaC79) or no TXNIP (HLaC79-Tax) induction could not be achieved by Galium aqueous extract.

Mouse double minute 2 homolog (MDM2) was significantly downregulated in Hep2 cells 48 h after Galium decoction treatment. Negative gene regulation by Galium aqueous extract was confirmed to be significant in Fadu cells, in HLaC79 cells without significance. The chemoresistant cell lines showed upregulation of MDM2 (significant in HLaC79-Tax, not significant in Hep2-Tax).

The regulation of RND3/RhoE, which belongs to the Rho family of small GTP-binding proteins was also examined. RND3 was downregulated significantly in Hep2, Fadu and HLaC79 cells following Galium verum treatment. Similar to MDM2, an inverse reaction in the chemoresistant clones (significantly upregulated in Hep2-Tax, not significant in HLaC79-Tax) was observed.