Mouse monoclonal antibody against ABCG2 BXP-21 (cat. no. ab3380) was purchased from Abcam (Cambridge, MA, USA). Allophycocyanin (APC)-conjugated anti-CD133/1 (cat. no. 130-090-826) antibody and rabbit anti-CD133 (cat. no. MAB4399) antibody were obtained from Miltenyi Biotec (Bergisch Gladbach, Germany) and EMD Millipore (Billerica, MA, USA), respectively. Rabbit monoclonal antibody directed against β-actin (cat. no. 4970) and horseradish-peroxidase-conjugated horse anti-rabbit and anti-mouse (cat. nos. 7074 and 7076) IgG secondary antibodies were purchased from Cell Signalling Technology, Inc. (Danvers, MA, USA). Fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG antibody (cat. no. #E031210-01) was purchased from EarthOx, LLC (San Francisco, CA, USA). Alexa Fluor^® 568-conjugated goat anti-rabbit IgG antibody (cat. no. #A-11011) was purchased from Invitrogen; Thermo Fisher Scientific, Inc. (Waltham, MA, USA). Reverse transcription (RT) and quantitative polymerase chain reaction (qPCR) kits were purchased from Takara Biotechnology Co., Ltd. (Dalian, China). Cisplatin was obtained from Sigma-Aldrich (St. Louis, MO, USA), and dissolved in a 0.15 M NaCl solution. Aliquots were stored at −20°C for up to 3 months and thawed immediately prior to use. All other chemicals and solvents were of the highest analytical grade available.

The A549 human LC cell line was obtained from American Type Culture Collection (Rockville, MD, USA) and cultured in RPMI-1640 medium (Hyclone; GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin at an atmosphere of 5% CO₂ at 37°C. Cells in the logarithmic phase of growth were used for all experiments.

Suspended cells were collected from confluent A549 cells and centrifuged at 800 × g for 5 min at room temperature. The pellets were gently dissociated with a pipette and resuspended in serum-free Dulbecco's modified Eagle's medium (DMEM)-F12 medium (Gibco; Thermo Fisher Scientific, Inc.). The cells were plated at a density of 1×10³ cells/ml in ultra-low attachment plates (Corning Inc. Acton, MA, USA) supplemented with 20 ng/ml basic fibroblast growth factor (bFGF; PeproTech, Inc., Rocky Hill, NJ, USA), 20 ng/ml epidermal growth factor (PeproTech, Inc.) and 2% B27 (Gibco; Thermo Fisher Scientific, Inc.) in a 5% CO₂ humidified incubator at 37°C. The medium was replaced every 2–3 days by carefully removing half the upper layer of medium and replacing it with an equal volume of fresh medium. It has been reported that the more serial passages in the spheres, the more CSCs present (4). The spheres were collected by gentle centrifugation at 600 × g for 6 min at room temperature, and mechanically dissociated into single cell suspensions until they were of a sufficient number (>200 cells) for passaging. The third passage of LC spheres was used in the present study to ensure the reliability of the results. The spheres were filtered using a cell strainer, and those with a diameter >40 μm were selected for use in the subsequent experiments. To determine the serum-induced differentiation of sphere-forming cells, spheroids were isolated and cultured in DMEM supplemented with 10% FBS without growth factors at 37°C in a 5% CO₂ humidified incubator.

The total RNA of the A549 attached cells and corresponding spheres were extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). First-strand cDNA was synthesized from 1 μg total RNA using M-MuLV Reverse Transcriptase (Applied Biosystems, Foster City, CA, USA) to a final volume of 20 μl. For RT-PCR, samples were assayed in a 12.5 μl reaction mixture containing 2.5 μl cDNA, 1.25 μl of 10× Ex Taq Buffer, 1 μl dNTP mixture, 1 μl forward primer, 1 μl reverse primer, 0.05 μl TaKaRa Ex Tag HS and 5.7 μl RNase Free dH₂O. The RT reaction was performed using a DNA Engine Peltier Thermal Cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The semi-quantitative PCR conditions were as follows: Initial denaturation at 95°C for 2 min, followed by 32 cycles of denaturation at 94°C for 30 sec, annealing at 55°C for 30 sec, and extension at 72°C for 30 sec, with extension at 72°C for 3 min in the final cycle. For qPCR, samples were assayed in a 10 μl reaction mixture containing 5 μl of 2× One Step SYBR RT-PCR Buffer, 0.45 μl of PrimeScript 1 Step Enzyme mix, 0.4 μl PCR forward primer (10 μM), 0.4 μl PCR reverse primer (10 μM), 2 μl cDNA and 1.8 μl RNase Free dH₂O using the One-Step SYBR^® PrimeScript™ qPCR kit (Takara Biotechnology Co., Ltd.). The qPCR was performed using an ABI 7500 Real-Time PCR system (Applied Biosystems) for a total of 40 cycles with an annealing temperature of 95°C for 5 sec and 60°C for 20 sec. The primer pairs encoding products of 113, 69, 126, 281, 206 and 225 bp, respectively, were as follows: CD133, forward 5′-GAGTCGGAAACTGGCAGATAGCA-3′ and reverse 5′-ACGCCTTGTCCTTGGTAGTGTTG-3′; ABCG2, forward 5′-CACAAGGAAACACCAATGGCT-3′ and reverse 5′-ACAGCTCCTTCAGTAAATGCCTTC-3′; octamer-binding transcription factor 4 (OCT4), forward 5′-ACATCAAAGCTCTGCAGAAAGAAC-3′ and reverse 5′-CTGAATACCTTCCCAAATAGAACCC-3′; Nanog, forward 5′-AGAAGGCCTCAGCACCTA-3′ and reverse 5′-GGCCTGATTGTTCCAGGATT-3′; sex-determining region Y-box 2 (Sox2), forward 5′-AGCAACGGCAGCTACAGCA-3′ and reverse 5′-TGGGAGGAAGAGGTAACCACAG-3′ and GAPDH, forward 5′-GAAGGTGAAGGTCGGAGTC-3′ and reverse 5′-GAAGATGGTGATGGGATTTC-3′. All primers were synthesized by BGI Tech Solutions Co., Ltd. (Beijing, China). The thermal cycling conditions were as follows: 95°C for 30 sec, followed by 5 sec at 95°C, and 1 min at 60°C for 40 cycles. Gene expression was normalized to GAPDH, and the fold change was calculated using the following formula: Fold change in gene expression = 2^−(ΔΔdCq) (21).

In total, 1.0×10⁶ viable attached cells and sphere-forming cells were collected. The cells were wasHed twice in phosphate-buffered saline (PBS), following which they were incubated with mouse monoclonal anti-ABCG2 (1:200 dilution) for 10 min at 4°C and then incubated with FITC-conjugated goat anti-mouse IgG (1:200 dilution) for another 10 min at 4°C in the dark. Subsequently, the cells were washed in PBS, resuspended, fixed in 100 μl 1% (m/v) paraformaldehyde and subjected to fluorescence analysis using flow cytometry. Dead cells were eliminated using 1 mg/ml propidium iodide (PI) staining (Sigma-Aldrich). To detect the expression of CD133, the cells were incubated with anti-CD133-APC for 15 min at 4°C in the dark, and subsequent procedures were performed, according to the protocol described above. The data were processed using CellQuest software (BD Biosciences, San Jose, CA, USA).

The A549 attached cells and sphere-forming cells were seeded on glass coverslips in at a density of 5.0×10³ cells/well six-well plates, and cultured in a 5% CO₂ incubator at 37°C overnight. The cells were fixed by immersion in ice-cold methanol for 15 min. Following washing of the coverslips twice in PBS and blocking for 30 min in blocking buffer (10% normal goat serum and 0.5% Triton X-100 in PBS), the coverslips were incubated with mouse anti-ABCG2 and rabbit anti-CD133 primary antibodies at a 1:200 dilution for 12 h at 4°C. The coverslips were then washed three times in PBS for 5 min each and incubated with FITC-conjugated goat anti-mouse IgG or Alexa Fluor^® 568-conjugated goat anti-rabbit IgG secondary antibody at a 1:200 dilution for 30 min at 37°C. Subsequently, the cells were counterstained with 4′,6-diamidino-2-phenylindole and visualized using an inverted fluorescence microscope (Leica, Mannheim, Germany).

The A549 adherent cells and sphere-derived cells were plated in triplicate at a density of 4,000 cells/well in standard coated 96-well plates and allowed to adhere overnight at 37°C in a 5% CO₂ humidified incubator. Wells containing medium only were used as a background negative control. Cell proliferation was measured during the following 4 days using a Cell Counting Kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan), according to the manufacturer's protocol. The spectrometric absorbance was measured at 450 nm using a microplate reader. The growth curves were constructed according to the mean value of absorbance relative to the background. To examine drug sensitivity, the cells were first seeded in a 96-well plate at a density of 5,000 cells/well and cultured in medium containing increasing concentrations of cisplatin (0–25.6 μg/ml) for 48 h at 37°C in a 5% CO₂ humidified incubator. At the end of the drug exposure duration, cell viability was measured using the CCK-8 assay, and the growth inhibition rate and the half maximal inhibitory concentration (IC₅₀) values were calculated using GraphPad Prism 5.0 software (GraphPad Software, Inc., La Jolla, CA, USA).

Cell invasion assays were performed using 24-well Transwell plates (8 μm pore size; Corning, Inc.) coated with 1 mg/ml Matrigel (BD Biosciences). In total, 1.0×10⁵ cells/well were suspended in 300 μl 0.1% bovine serum albumin (Invitrogen; Thermo Fisher Scientific, Inc.)/serum-free medium and added to the upper compartment of the Transwell plates. Subsequently, 500 μl complete medium containing 10% FBS was added to the lower wells of each plate, and the cells were incubated in a 5% CO₂ incubator for 24 h at 37°C. After 24 h, the upper surface of the chambers were scraped using cotton swabs, and the cells on the lower surface of the membrane were considered invasive cells. These invasive cells were fixed for 15 min in methanol, stained with Giemsa solution and examined using a TE2000 inverted micro scope (Nikon, Tokyo, Japan).

For cell cycle analysis, the adherent A549 cells, spheres and sphere-forming cells following serum-induced culture were harvested by trypsinization and fixed with ice-cold 70% ethanol overnight at 4°C. The cells were then washed twice in PBS, resuspended in PBS and stained with PI/RNase staining buffer (BD Biosciences) at 37°C for 30 min. The cell cycle profiles were obtained using flow cytometry at 488 nm, and the data were analyzed using CellQuest software (BD Biosciences).

Cell lysates of A549 attached cells and spheres were prepared according to manufacturer's details of the protein extract kit (KEYGEN Biotech, Beijing, China). Protein concentrations were determined using the bicinchoninic acid method. Equal quantities of protein (50 μg/lane of total extract) were separated by 12% sodium dodecyl sulphate-polyacrylamide gel electrophoresis and then electrophoretically transferred onto PVDF membranes (Bio-Rad Laboratories, Inc.). The membranes were blocked in 5% skim milk powder dissolved in Tris-buffered saline with Tween 20 (TBST) for 2 h at room temperature, following which they were incubated with the following primary antibodies at 4°C overnight: ABCG2, CD133 and β-actin. The membranes were then washed three times in TBST buffer, and were subsequently incubated with secondary antibody for 1.5 h at room temperature. The bands were visualized using a molecular imaging system. Densitometric analysis was performed using Image-ProPlus software (Media Cybernetics, Inc., Rockville MD, USA), and β-actin was used as a control.

Male (n=20) and female (n=200 athymic BALB/c nude mice (4–6 weeks old) with body weights of 18–22 g were provided by the Experimental Animal Centre of Chongqing Medical University (Chongqing, China). All mice were housed in microinsolator cages in a specific homothermic pathogen-free environment with a 12-h light/dark cycle and access to food and water ad libitum. To generate subcutaneous xenograft models, different numbers (1×10⁶, 2×10⁵, 1×10⁵, 4×10⁴ and 1×10⁴) of freshly dissociated attached cells and sphere-derived cells were mixed with 100 μl PBS and subcutaneously inoculated into the left and right flank of each BALB/c nude mouse to detect their tumorigenic capacity. The tumorigenicity was measured primarily by tumor size, tumor weight and latency, the latter determined as the duration between injection and the detection of palpable tumors. The primary tumor sizes were measured with a calliper on a weekly basis, and approximate tumor weights were determined using the following formula: V = 1/2 × a × b², where b is the smaller of the two perpendicular diameters and V is tumor volume. When the average tumor size reached 1,000–2,000 mm³, the animals were sacrificed by cervical dislocation. The grafts were then removed and fixed in formalin, and paraffin sections were prepared for hematoxylin and eosin (H&E) staining and analyzed under a light microscope. All of the in vivo experimental procedures were approved by the Committee on the Ethical Use of Animals of the First Affiliated Hospital of Chongqing Medical University.

All experiments were performed in triplicate. The data are graphically represented as the mean ± standard error of the mean. The values were compared with controls using either Student's t-test or two-way analysis of variance using Prism GraphPad 5.0 software (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference.

The A549 attached cells were cultured in serum-free medium, as described above. Under ultra-low attached conditions, the cells grew into floating, tridimensional clusters, termed spheres. The spheres began to form on day 3 or 4, and formation was more substantial on day 7. Between days 9 and 12, the spheres had completely formed. By days 15–16, the spheres had become well-rounded structures, which were composed of abundant, cohesive cells. Fig. 1A shows the generation of a sphere from a single LC A549 cell. After 48 h of serum-driven culture, floating spheres were found to have adhered and migrated to the wall of the culture flasks, and gradually differentiated into adherent cells.

To examine the subcellular localization of CD133 and ABCG2 in the sphere-forming cells, immunofluorescence staining of CD133 and ABCG2 was performed. Positive staining was observed, with CD133 and ABCG2 present primarily in the membranes of the spheres. Dual staining for CD133 and ABCG2 indicated that the two candidate CSC markers were colocalized in the spheres (Fig. 1B).

To clarify the differential gene expression profiles between the spheres and the adherent A549 cells, RT-PCR and qPCR analyses were performed (Fig. 2A and B). As expected, cancer cells cultured in serum-free medium caused a shift in CSC markers, including marked upregulation of the following CSC markers: CD133, ABCG2, Oct-4, Sox-2 and Nanog. Compared with the attached cells, the flow cytometry data (Fig. 2C) showed significantly increased expression levels of CD133 (0.35±0.06, vs. 13.37±1.66%) and ABCG2 (20.63±13.18, vs. 73.72±11.57%) on the spheres.

The results of the flow cytometric analysis revealed the existence of distinct subpopulations with differential expression of CD133 and ABCG2 in the tumorospheres; a small subpopulation of CD133⁺/ABGC2⁺ cells was identified in the A549 cells, whereas the primary cell population was negative for the two CSC markers. To investigate possible functional differences between these subpopulations, the CD133⁺/ABCG2⁺ cells were subjected to FACS (Fig. 3A) and assessed using several functional assays. The CCK-8 assays revealed that the A549 spheres had markedly higher proliferation potential, as shown in Fig. 3B. Consistently, the results from the colony formation assay suggested that the CD133⁺/ABCG2⁺ sphere-derived cells had higher cloning efficiency (3–4-fold; 368.569±70.96, vs. 100±17.321) leading to the formation of larger colonies at earlier time points, with more cells having the ability to initiate a clone (Fig. 3C). In addition, the sphere-derived cells appeared to penetrate the Matrigel more readily (Fig. 3D), and the number of stained invasive cells was greater, compared with the attached cells (227.293±66.781, vs. 100±8.66). Therefore, the LC tumorospheres were found to possess augmented proliferation and invasion abilities.

It is well known that there is a close association between cell cycle and cell function, including the ability of proliferation and differentiation, and chemosensitivity (22). In order to clarify the difference of cell cycle distribution between the A549 adherent cells and spheres, the cell cycle was analyzed using flow cytometry. The results (Fig. 3E) revealed that the A549 spheres formed in serum-free conditions exhibited the following cell cycle distribution: 76.69±4.31% in the G0/G1 phase, 14.02±2.65% in The S phase and 9.29±2.70% in the G2/M phase. However, the A549 attached cells exhibited the following cell cycle distribution: 55.61±5.88% in the G0/G1 phase, 32.38±6.24% in the S phase and 12.01±2.47% in the G2/M phase. Therefore, the cell cycle of the spheres was arrested in the G0/G1 phase, with a corresponding reduction in the numbers of cells in the S and G2/M phases. However, in response to the serum stimulus, the spheres acquired the following cell cycle phase distribution: 16.77±8.92% in the G0/G1 phase, 73.64±10.06% in the S phase and 9.59±5.04% in the G2/M phase. This distribution indicated that the A549 spheres were steadily maintained in a quiescent state in serum-free conditions, which may partly contribute to the refractoriness to cancer therapy. However, the cell cycle progress was promoted with a significantly increased ratio of sub S cells following treatment with serum, suggesting that the spheres had the capacity for unlimited self-renewal and extensive proliferation, and to generate differentiated progeny.

To examine whether LC tumorospheres possess the hypothesized chemoresistant phenotype of CSCs, the present study assessed the sensi tivity of the sphere-forming cells and differentiated cells to cisplatin, which is commonly used in chemotherapy. The LC sphere-forming cells exhibited an increased IC₅₀ value (11.752±1.937, vs. 1.485±0.121 μg/ml), as shown in Fig. 4A. In addition, treatment with 5 μg/ml cisplatin did not affect the number or the size of the tumorospheres, suggesting enhanced chemoresistance of the sphere-forming cells (data not shown). The chemoresistance of tumor cells occurs, in part, due to the overexpression of ATP-binding cassette family members, including ABCG2, P-glycoprotein and multidrug resistance protein 1 (23,24). CD133 has also been reported to be partially associated to multidrug resistance. In the present study, the expression levels of CD133 and ABCG2 were further confirmed at the protein level by immunoblotting. The data indicated that the protein expression levels of CD133 and ABCG2 were substantially increased in tumorospheres, compared with adherent cells (Fig. 4B).

The cancer stem-like properties were further confirmed in the LC spheres through in vivo tumorigenicity experiments. The data showed that the A549 sphere-derived cells generated subcutaneous tumors with larger volumes in a shorter duration, compared with those from the attached cells. The tumor growth curves were compared for the two groups (Fig. 4C and D). In addition, as few as 4×10⁴ cells from the A549 sphere-forming cells were able to form tumors at earlier time points when subcutaneously injected into nude mice, whereas 2×10⁶ attached cells were required for the same effect; this value was 50-fold higher than that for the sphere-forming cells. No differences in the overall health or conditions of vital organs were observed between the mice in the two groups (data not shown). The xenografts were confirmed as lung adenocarcinoma using H&E staining (Fig. 4E), and the transplanted tumors derived from the A549 spheroid-forming cells were histologically similar to those from A549 adherent cells, predominantly containing differentiated cells. These data suggested that the LC sphere-derived cells showed an increased capacity for xenograft tumor formation in vivo.