Cell culture media, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), hematoxylin and eosin (H&E) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Fetal bovine serum (FBS) and antibiotics were from Life Technologies (Carlsbad, CA, USA) and Immobilon P (PVDF) membranes from Millipore (Billerica, MA, USA). OOS was provided by Catalysis, S.L. (Madrid, Spain). Its formulation included per 50 ml of solution: L-glycine 1,000 mg, glucosamine sulfate 1,000 mg, L-arginine 320 mg, L-cysteine 102 mg, licorice (Glycyrrhiza glabra L.) 100 mg, vitamin C (L-ascorbic acid) 60 mg, water, zinc sulfate 40 mg, green tea extract (Camellia sinensis L. Kuntze) 12.5 mg, vitamin B5 (D-calcium pantothenate) 6 mg, vitamin B6 (pyridoxine hydrochloride) 2 mg, manganese sulfate 2 mg, cinnamon extract (Cinnamomum zeylanicum Blume) 1.5 mg, folic acid (pteroylmonoglutamic acid) 200 μg, vitamin B12 (cyanocobalamin) 1 μg, acidulant (malic acid) and preservative (sodium methylparaben). Other generic chemicals were purchased from Sigma-Aldrich, Roche Biochemicals (Basel, Switzerland) or Merck (Darmstadt, Germany).

The origins of the different antibodies used in the western blot analyses were as follows: the anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (sc-166574, mouse monoclonal, used at 1:10,000) and the anti-PARP (sc-8007, mouse monoclonal, 1:5,000) antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA); the anti-Rb (#554136, mouse monoclonal, 1:1,000) and anti-caspase-3 (#610323, rabbit polyclonal, 1:5,000) from BD Biosciences (Franklin Lakes, NJ, USA) and the anti-p27 (#3686, rabbit polyclonal, 1:5,000), anti-caspase-8 (#9746, mouse monoclonal, 1:1,000), anti-caspase-9 (#9502, rabbit polyclonal, 1:1,000), anti-caspase-7 (#9492, rabbit polyclonal, 1:1,000), anti-cleaved caspase-3 (#9664, rabbit polyclonal, 1:1,000) and anti-bid (#2002, rabbit polyclonal, 1:1,000) from Cell Signaling Technology, Inc. (Danvers, MA, USA). The horseradish peroxidase-conjugated secondary antibodies (goat anti-rabbit, 170-6515, 1:20,000 and goat anti-mouse 170-6516, 1:10,000) were from Bio-Rad Laboratories Inc. (Hercules, CA, USA).

SCLC cell lines were grown in RPMI-1640 medium supplemented with 10% FBS and antibiotics. Cells were cultured at 37°C in a humidified atmosphere in the presence of 5% CO₂-95% air. The two SCLC cell lines used in this study have already been described (21–23) and were generously provided by Dr Clementi (CNR Center of Cytopharmacology, Milan, Italy). Furthermore, GLC-8 cells were initially isolated from a tumor biopsy of an SCLC (22), while DMS 92 cells came from a bone marrow metastasis of an SCLC (21).

To prepare for protein purification, cells were harvested by centrifugation, washed in phosphate-buffered saline (PBS) and lysed in ice-cold lysis buffer (24). Protein quantification, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and WB analysis were performed as previously described (24).

For cell cycle analysis by flow cytometry, ethanol-fixed cells were stained with 5 μg/ml propidium iodide (PI) and 250 μg DNase-free RNAse. A total of 50,000 cells were acquired in the PI gate by using a BD Accuri™ C6 flow cytometer and the C6 (version 1.0.264.21) software (BD Biosciences).

To measure bromodeoxyuridine (BrDU) incorporation, the Cell Proliferation ELISA, BrdU (colorimetric) kit (Roche Diagnostics GmbH, Mannheim, Germany) was used following the manufacturer's instructions. Data were acquired using a BioTek Synergy 4 reader with Gen5 1.05 software (both from BioTek Instruments, Inc., Winooski, VT, USA).

For analysis of apoptotic cell death, GLC-8 cells were treated for 48 h with OOS diluted 1:25 in culture media or left untreated (control), resuspended in binding buffer (10 mM HEPES, 140 mM NaCl, 2.5 nM CaCl₂, pH 7.4) containing 5 μl Annexin V-FITC (BD Biosciences) and 5 μl 50 μg/ml PI, and stained at room temperature for 15 min. A total of 50,000 cells were acquired using the BD Accuri™ C6 flow cytometer (BD Biosciences) as previously described.

Cell proliferation of SCLC cells was examined using a modified MTT metabolization assay (19). At least three wells were analyzed for each condition, and the results are presented as the mean ± SD of a representative experiment repeated at least twice.

For the animal studies, 12 7-week-old female athymic mice (CB17-SCID) were purchased from Charles River Laboratories (Wilmington, MA, USA) and maintained in pathogen-free housing at our Institutional Animal Care Facility. Animal experiments were performed according to the institutional guidelines and protocol approved by the Ethics Committee of CSIC-Universidad de Salamanca (Salamanca, Spain). One week later, 6×10⁶ GLC-8 cells resuspended in 50 μl of RPMI-1640 and 50 μl of Matrigel were subcutaneously injected into the right caudal flank of each animal. When tumors became palpable, mice were randomized into two groups (n=6 per group), that were administered vehicle alone (control group) or 100 μl OOS per mouse (20 g). Treatments were administered for 31 days with a daily schedule (Monday to Friday) by oral gavage. Mice were weighed and tumors measured twice a week with a digital caliper (Proinsa, Vitoria, Spain). Tumor volumes were calculated using the following formula: V = (L/2) × (W/2)² × 4/3 × π, where V, volume (cubic mm); L, length (mm); and W, width (mm). At the time of sacrifice, tumor tissue was resected and part of the tumor was immediately frozen at −80°C. Another part was fixed in formalin for further analyses.

Representative tumor areas were fixed in formalin, paraffin-embedded, cut in 2- to 3-μm sections, and either stained with H&E or prepared for IHC, that was performed as previously described (19). Thus, two cell conditioning periods of 8 min at 95°C and 4 min at 100°C on a hot plate using Tris-EDTA (pH 8.0) buffer were performed on previously dewaxed formalin-fixed paraffin-embedded sections. Sections were then incubated with the a 1:50 dilution of an anti-Ki-67 antibody (clon SP6; Master Diagnóstica, Granada, Spain) and the staining was performed with the IHC 3,3′-diaminobenzidine (DAB) system (Ventana Medical Systems, Tucson, AZ, USA). For TUNEL staining, the In Situ Cell Death Detection kit (Roche Diagnostics GmbH, Mannheim, Germany) was used, and counterstained with 4′,6-diamino-2-phenylindole. Results were evaluated in a manner blinded to the clinicopathological and molecular data. The number and intensity of immunoreactive cells were evaluated in at least 10 randomly selected fields. These procedures were conducted by independent personnel of the pathology unit of our center. Conflict measurements were solved by consensus.

After thawing, tumors were excised and lysed in TRIzol reagent (Life Technologies), according to the manufacturer's instructions. Briefly, tumors were homogenized (Dispomix; L&M Biotech, Holly Springs, NC, USA) and incubated in TRIzol solution for 2 min at room temperature, before the addition of chloroform. Tubes were vigorously shaken and the different phases were separated by centrifugation at 18,000 × g and 4°C for 15 min. The upper, aqueous phase was recovered and the RNA present was precipitated with isopropyl alcohol. Once washed in 70% ethanol, the resultant RNA was column-purified (RNeasy Mini kit; Qiagen, Inc., Valencia, CA, USA) and its integrity was assessed (Agilent 2100 Bioanalyzer; Agilent Technologies, Inc., Santa Clara, CA, USA). Biotinylated complementary RNA was then synthesized (Enzo Life Sciences, Inc., Farmingdale, NY, USA) and hybridized to human Clarion™ S GeneChip oligonucleotide arrays (Affymetrix; Thermo Fisher Scientific, Inc.). Quantitation of fluorescence intensities of probesets was conducted using the GenArray Scanner (Hewlett Packard). Unprocessed files were normalized using the RMA algorithm implemented in the Affymetrix Expression Console (Thermo Fisher Scientific, Inc.). Differentially expressed genes were identified using significant analysis of microarrays, selecting all genes with a value of Q≤0.05.

Each condition was analyzed in triplicate or quadruplicate and data are presented as the mean ± SD of an experiment that was repeated at least three times. Comparisons of continuous variables between two groups were performed using two-sided Student's t-test. Differences were considered to be statistically significant when P-values were <0.05.

The in vitro action of OOS was evaluated in the SCLC cell lines GLC-8 and DMS 92. With this purpose, cells were grown in the presence of increasing doses of OOS diluted in the culture medium and their proliferation was evaluated using MTT metabolization assays performed at different days of treatment. OOS decreased MTT metabolization in these cells in a time- and dose-dependent manner (Fig. 1A and C). The IC₅₀ values determined for GLC-8 cells after 4 days and DMS 92 cells after 6 days of treatment were a 1:86 dilution and a 1:60 dilution, respectively (Fig. 1B and D).

In most of the cases, the success of antitumor therapies is based on the combination of different agents. For this reason, we aimed to ascertain whether the combination of OOS with drugs commonly used in the SCLC clinic could improve their antitumor effect. Thus, GLC-8 cells were treated for 2 days with OOS alone (diluted 1:100 in regular media), or in combination with vincristine (VCR, 5 nM), docetaxel (TXT, 5 nM) or cisplatin (2 or 20 μM). The combination of OOS with this latter compound did not exhibit a superior antiproliferative effect to the one observed for each of the individual treatments (data not shown). In the case of VCR or TXT, the combinations were more effective for inhibiting MTT metabolization than the independent treatment with any of the agents (Fig. 1E and F). In fact, when the combination index (CI) of OOS and VCR was calculated using Calcusyn software (Biosoft, Cambridge, UK) an additive effect of both drugs was demonstrated (CI between 0.9 and 0.96, data not shown).

Given the antitumoral action of OOS in vitro, we explored whether this effect could also be observed in vivo. With this aim, we used a xenograft murine model to determine the effect of OOS treatment once the tumors were already established. CB17-SCID athymic mice were injected with GLC-8 cells and when tumors were palpable and had correctly been engrafted and started to grow, the animals were randomized into two different groups that were orally treated with water (control) or 100 μl OOS per 20 g of animal weight (OOS). Initially, tumors included in the two groups had a similar volume (209.3±17.2 vs. 206.1±16 mm³, mean ± SEM, respectively). Tumoral sizes were measured twice a week for a total time period of 31 days. After 14 days of oral treatment with OOS, a significant decrease in the growth of the tumors was observed (p=0.027) (Fig. 2A). Nonetheless, such differences increased along the time of treatment. Thus, at the end of the experiment, the mean tumor volume of the control mice was 2,402.6±223.6 mm³, as compared with 1,197±144.4 mm³ for the group treated with OOS (p=0.003) (Fig. 2A). The mean body weight of the control or treated mice did not substantially change throughout the duration of the experiment (Fig. 2B), indicative of good tolerability to OOS in our experimental conditions.

The mechanisms leading to the in vitro reduction of MTT metabolization were next analyzed. We proposed that such a reduction could be due to an increase in cell death, reduced cell cycle progression or a combination of both. Thus, the capability of OOS to induce apoptotic cell death was assessed. GLC-8 cells were treated for 24 h with OOS diluted at 1:25 in culture medium and apoptotic cell death was determined by double staining with Annexin V and PI. A slight increase in the percentage of cells stained by both Annexin V and PI was observed after OOS treatment (control, 16.7% vs. OOS-treated, 32.0%) (Fig. 3A). Analysis of the total viable vs. non-viable population confirmed the induction of cell death by OOS (Fig. 3B). Moreover, when the levels of the cleavage of several proteins involved in apoptosis such as PARP or different caspases were determined by western blotting, a clear induction of their proteolytic processing indicative of activation was detected, especially at dilutions of OOS of 1:25 (Fig. 3C).

We next investigated whether OOS could induce changes in cell cycle progression. Cells were treated with OOS and cytometrically analyzed after PI staining of the DNA. In those conditions, treatment with OOS resulted in an increase in the G0/G1 population, increasing from 74.8 to 88.5%. Consequently, there was a concomitant decrease in the percentage of cells in the G2/M (from 14.1 to 7.2%) or S phases (from 11.1 to 4.3%) following treatment with OOS (Fig. 3D). The capability of OOS to interfere with DNA synthesis was evaluated using an alternative technique that allows the quantification of cells progressing through S phase by the measure of its BrDU incorporation. With this assay, a reduction in the S phase population was demonstrated (Fig. 3E), pointing to a slowdown in cell cycle progression as part of the mechanism of action of OOS.

The cell cycle effects of OOS in other cellular models seems to be mediated by an increase in the cell cycle inhibitory protein p27 (13,19). To explore whether such a mechanism of action could as well participate in the effect of OOS on SCLC cells, p27 protein levels were determined by western blotting both in control untreated cells or after 24 h of treatment with 1:25 or 1:50 dilutions of OOS. When a dilution factor of 1:25 was used, a clear increase in the p27 level was observed (Fig. 3F). Moreover, this increase was correlated with a decrease in the phosphorylation and total levels of the retinoblastoma protein (pRB and RB, respectively) (Fig. 3F).

We next investigated the potential mechanisms leading to the decrease in tumor growth observed in OOS-treated animals, when compared to the untreated controls. We first evaluated the general tumor aspects after H&E staining. Independently of whether the animal received treatment of not, the tumoral masses dissected from the animals presented a large central apoptotic/necrotic area that extended to up to 60% of the tumor (data not shown). Tumor cells were in general poorly differentiated, with a high mitotic index. Additionally, control tumors seemed to be better vascularized than OOS-treated ones. In fact, control tumors seemed to be more pseudo-glandular than the treated ones (arrows in Fig. 4A). In line with this, tumors from animals which received OOS exhibited a more compact aspect (Fig. 4A).

We next evaluated the level of apoptotic cells in the control and OOS-treated animals. While, as mentioned above, when observed under a microscope the central apoptotic/necrotic area observed was similar in both experimental conditions, in OOS-treated tumors more apoptotic regions within non-necrotic tissue could be observed. Analysis of apoptotic cell death after TUNEL staining technique showed that tumors from the OOS-treated animals contained a larger number of apoptotic cells than tumors from untreated animals (Fig. 4B). Moreover, quantitative analyses of the TUNEL-stained tumors corroborated that the number of apoptotic cells detected per field was significantly lower in the control condition than that in the treated tumors (p=0.019) (Fig. 4C). Together, these data demonstrate that OOS triggers the apoptotic cell death of SCLC cells in the in vivo setting.

The expression of human Ki-67, a protein associated with cell proliferation, was also evaluated. Treatment with OOS induced a statistically significant decrease in the percentage of Ki-67-positive cells (Fig. 4D and E). The decrease in Ki-67 levels was indicative of a reduced proliferation of OOS-treated tumors.

To gain further insight into the molecular mechanism leading to the in vivo reduction in tumor growth, we investigated the changes in gene expression profiles induced by OOS treatment. Unsupervised clustering of the gene expression data clearly associated samples into two groups, control and OOS-treated, indicating that treatment with OOS was able to change the transcriptome to make treated samples to diverge from the untreated control tumors (Fig. 5A). Quantitative comparison of both conditions defined 68 deregulated genes (fold change of at least ±1.5) (Table I). Volcano plots indicated that among them, 31 genes were upregulated in the control condition whereas 37 were increased after OOS treatment (Fig. 5B). Among the deregulated genes, NADH dehydrogenase NDUFA4L2 (2.26-fold higher expression), the ID1 inhibitor of DNA binding (2.2-fold) or ITGB2 (1.7-fold) had a higher expression in the control samples than in those from animals treated with OOS. In contrast, the opposite situation was the olfactory receptor OR51F2 (2.17-fold higher expression in OOS-treated samples) or YBX1 acceptor (2.12-fold), among others (Table I). Additionally, several intracellular pathways were deregulated, including the DNA damage response pathways, cell cycle regulation, or the PI3K/AKT/mTOR, Wnt or insulin signaling, all of them involved in cell proliferation/apoptosis (Fig. 5C). Moreover, when the GSEA software was used to analyze these gene-expression data, the p27 pathway (Fig. 5D) and certain pro-apoptotic pathways (data not shown) were found to be upregulated in the OOS-treated tumors. In addition, several angiogenesis-related datasets were downregulated in tumors derived from OOS-treated animals, as shown in Fig. 5E.