Cranberry Proanthocyanidins Mediate Growth Arrest of Lung Cancer Cells through Modulation of Gene Expression and Rapid Induction of Apoptosis

Cranberries are rich in bioactive constituents purported to enhance immune function, improve urinary tract health, reduce cardiovascular disease and more recently, inhibit cancer in preclinical models. However, identification of the cranberry constituents with the strongest cancer inhibitory potential and the mechanism associated with cancer inhibition by cranberries remains to be elucidated. This study investigated the ability of a proanthocyanidin rich cranberry fraction (PAC) to alter gene expression, induce apoptosis and impact the cell cycle machinery of human NCI-H460 lung cancer cells. Lung cancer is the leading cause of cancer-related deaths in the United States and five year survival rates remain poor at 16%. Thus, assessing potential inhibitors of lung cancer-linked signaling pathways is an active area of investigation.


Introduction
Lung cancer is the leading cause of cancer related death among men and women in the United States and despite recent advances in treatment overall prognosis remains poor [1]. The development of effective agents for the prevention and treatment of lung cancer is an active area of investigation.
Smoking is the principal cause of lung cancer [2]; however, risk increases with exposure to second hand smoke, radon, asbestos, radiation, arsenic, aluminum, chromium, cadmium and select organic chemicals [3]. The potential protective role of diet in lung cancer is still being unraveled. To date, epidemiological studies strongly support that diets rich in fruits may reduce lung cancer risk, as extensively reviewed in the World Cancer Research Fund report [4]. Fruits contain a multitude of bioactive food constituents with pleiotropic health benefits. Cranberries (Vaccinium macrocarpon Ait.), for example, reportedly have antimicrobial, antiviral and more recently, anticancer functions [5][6][7][8][9][10][11][12][13][14][15][16][17]. The current study sought to investigate potential cancer inhibitory mechanisms associated with a proanthocyanidin rich cranberry fraction (PAC) in "resistant" lung cancer cells. The investigation of such food-derived agents holds particular promise given the repeated negative results of vitamin/mineral/antioxidant supplementation trials in at risk cohorts [18][19][20][21]. Promising food-derived agents with cancer inhibitory effects supplied at behaviorally achievable levels are likely to be well tolerated and safe. These are important considerations for relatively healthy cohorts who may consume cancer protective agents for extended periods to derive maximum health benefits. Understanding the specific mechanisms of cancer inhibition is important not only for preventive interventions, but may hold promise for reversing a "resistant" phenotype prior to cancer chemotherapy.

Modulation of global gene expression patterns by PAC
To explore the mechanisms and signaling cascades linked to the cancer inhibitory potential of PAC we utilized global gene expression analysis of NCI-H460 human lung cancer cells which were treated with 50 µg/mL PAC or vehicle for 6 hours. This concentration of PAC was chosen based on earlier work which had determined the IC 50 to be 50 µg/mL [17]. Ease analysis was employed to investigate the effects of PAC on over-represented Gene Ontology (GO) categories ( Table 1). The top 30 biological processes down-or up-regulated greater than 2-fold are displayed in Table 1. As shown a large number of processes linked to cell death were significantly up-regulated by PAC treatment including regulation of apoptosis, regulation of programmed cell death, positive regulation of cell death, positive regulation of apoptosis, and apoptotic mitochondrial changes. The dominate biological processes down-regulated following PAC treatment included metabolism, protein modifications, and cell cycle linked processes as illustrated by down regulation of M phase of mitotic cell cycle, mitosis, M phase, cell cycle checkpoint, and regulation of cell cycle process. Figure 1 illustrates the cell cycle pathway in KEGG with red stars marking down-regulated markers following PAC treatment of NCI-H460 cells compared to vehicle treated NCI-H460 cells. A number of cell cycle related markers were validated utilizing real time PCR and RT² Profiler PCR Arrays including, BCCIP, CCNB1, CCND1,  CCNT1, CDC2, CDC16, CDC20, CDK4, CDK6, CDKN3, CHEK2, GTF2H1, HUS1, KNTC1,  MAD2L1, MCM5, MKI67, MNAT, MRE11A, PCNA, RAD1, RAD17, SERTAD1, SUMO1 and  TFDP1. Further evaluation of signaling pathways utilizing PANTHER further supported the general alterations noted on gene ontology categories. Detoxification processes were up-regulated by PAC treatment, as a number of glutathione S-tranferases were significantly increased. PAC treatment also down-regulated DNA metabolism, cell cycle, meiosis, mitosis and interestingly oncogenesis signaling. In addition, analysis of KEGG pathways following PAC treatment revealed effects on erbB and mTor signaling, pathways altered in lung carcinogenesis [22,23].   [24,25]. Red stars mark genes down regulated following PAC treatment of NCI-H460 cells compared to vehicle treated cells based on global gene expression results. As illustrated, PAC alters expression of a large number of genes involved in all phases of cell cycle regulation, including the G1/S and G2/M transitions and DNA replication.

Pac induces rapid and significant apoptosis in lung cancer cells
Next, PACs cell death inducing effects were further investigated over time in lung cancer cells via Annexin V-FITC staining methods and flow cytometry. As Figure 2 shows, PAC treatment resulted in significant and rapid apoptosis induction, with maximal apoptosis of 6.29-fold occurring 6 hours posttreatment. Total apoptosis occurred at 25.2, 55.7, 61.2, 55.6 and 21.4% at 2, 6, 12 and 24 hours, respectively. Specifically PAC induced early, late and total apoptosis following 2, 6, 12 and 24 hours of PAC treatment; however, by 48 hours there was a significant, but only modest 2.5 fold increase in total apoptosis supporting that the cells are starting to recovery from PACs cell death inducing effects at this late time-point. At 2 hours post-PAC treatment the majority of apoptosis is early, but rapidly shifts to late apoptosis by 6 hours. The data also shows significant reductions in "unstained" or live cells as displayed in the lower left quadrant of Figure 2B and 2C. The reductions in unstained vehicle versus PAC treated cells were 18.11, 48.19, 57.32, 48.91 and 13.74% at 2, 6, 12, 24, and 48 hours, respectively supporting that the greatest PAC-induced cell death occurs between 6 and 24 hours. The gene expression results coupled with these findings led us to further explore and validate specific apoptotic markers utilizing real time PCR as described in the methods section. Table 2 summarizes the results of the validated apoptotic markers and their potential functions.  In summary, PAC treatment resulted in rapid and significant induction of cell death in NCI-H460 lung cancer cells. This line is known to express wild type p53 [26], but also has been documented to over-express X-linked inhibitor of apoptosis protein (XIAP) resulting in suppressed activation of downstream effector caspases and apoptotic resistance [27]. Resistance is problematic in the context of A cancer prevention as well as chemotherapy; thus, it is particularly promising that PAC has potent cell death inducing effects in a "resistant" cell line. Furthermore, global gene expression results showed that PAC down-regulated expression of a number of inhibitor of apoptosis proteins (IAPs) including BIRC1, BIRC2, BIRC4 or XIAP, and BIRC6, many of which have been been linked to apoptosis resistance in the presence of anticancer drugs [28,29]. As shown in Table 2, PAC down-regulates a number of additional anti-apoptotic molecules, including BAG4, BNIP2, and BNIP3L. Conversely, the RT² Profiler PCR Apoptosis Array validated a number of pro-apoptotic markers [31,32] as upregulated following PAC treatment including BID, multiple pro-apoptotic TNF superfamily members and related adapter molecules, as well as p73 [33]. P73 is a p53 family member linked to cell death induction via apoptosis and type II cell death or autophagy [34]. Kim et al. reported that up-regulation of autophagy occurred following treatment with inhibitors of caspase-3 and mTOR resulting in enhanced radiosensitivity in a mouse model of lung cancer [35]. PAC treatment of NCI-H460 cell decreased pro-apoptotic, CASP3 [36] which may be linked to activation of type two cell death. PAC treatment up-regulated pro-apoptotic Bcl-2 family members [37] such as Bok, Bax, and Bad; the latter two molecules are known to be inactivated by tobacco specific carcinogens in lung epithelial cells [38][39][40][41]. Although further analysis of specific caspase molecules is warranted, the gene expression results coupled with the real-time PCR validation data support that apoptosis induction is mediated in part by the activation of death receptors belonging to the tumor necrosis factor receptor gene superfamily.   Next, a limited number of specific proteins were analyzed to further verify the effects of PAC on the apoptotic marker PARP1 and P21, an important mediator of cell cycle arrest as illustrated in Figure  3. We have previously reported that PAC decreases S-phase and causes arrest of NCI-H460 cells at the G1 checkpoint; however, we have recently found that PAC can induce arrest at the G2 checkpoint in esophageal adenocarcinoma cells (EAC) (unpublished data) supporting that specific cell death inducing effects differ between cell lines, likely due to the molecular profile of the individual cell line under investigation. P21 is reported to mediate cell cycle arrest in response to the p53 checkpoint [60] and the NCI-H460 cell line expresses wild type p53; whereas p53 is mutated or deleted in our EAC cells. In addition, induced activation of p21 has been linked to lung cancer cell growth inhibition and enhanced chemosensitivity to cisplatin [61].

Cell cultures
The lung adenocarcinoma cell line NCI-H460 originated from a non-small cell lung cancer of the large cell type and is available through American Type Culture Collection [62]. Cancer cells were grown in Dulbecco's modification of Eagle's medium (DMEM) containing L-glutamine (2.0 mM), penicillin (10 4 units/mL), sodium pyruvate (1 mM), and FBS (0-10%, depending on the experiment). Cells were maintained as monolayers (37 °C, 5% CO 2, 95% air).

Cranberry proanthocyanidins
Cranberry fruit (Vaccinium macrocarpon Ait.) was collected at the Marcucci Center for Blueberry and Cranberry Research, Chatsworth, NJ, USA. The cranberry PAC-rich powder was prepared Dr. Amy Howell (Rutgers University, Chatsworth, NJ) as previously reported [63]. In brief, purified cranberry proanthocyanidin extract was isolated from cranberries of the 'Early Black' cultivar using solid-phase chromatography according to well-established methods [63][64][65]. The fruit was homogenized with 70% aqueous acetone, filtered, and the pulp was discarded. Acetone was removed and the cranberry extract was suspended in water, applied to a preconditioned C-18 solid phase chromatography column, and washed with water to remove sugars, followed by acidified aqueous methanol to remove acids. The fats and waxes were retained on the C-18 sorbent. The total polyphenolic fraction containing anthocyanins and flavonol glycosides as well as the proanthocyanidins (confirmed using reverse phase HPLC with diode array detection) was eluted with 100% methanol and dried under reduced pressure. The total polyphenolic fraction was suspended in 50% EtOH and applied to a preconditioned Sephadex LH-20 column, which was washed with 50% EtOH to remove low molecular weight phenolics (anthocyanins and flavonol glycosides). Remaining proanthocyanidins that adsorbed to the LH-20 column were eluted with 70% aqueous acetone. Elution of the proanthocyanidins was monitored using diode array detection at 280 nm. The absence of absorption at 360 and 450 nm confirmed that anthocyanins and flavonol glycosides were successfully removed. Acetone was removed under reduced pressure, and the resulting purified proanthocyanidin extract was freeze-dried. The presence of proanthocyanidins with A-type linkages was confirmed using matrix-assisted laser desorption ionization (MALDI-TOF MS) or electrospray ionization (DI/ESI MS) as previously described [61]. To summarize, current technologies including 13 C NMR, electrospray mass spectrometry, MALDI-TOF MS, and acid-catalyzed degradation with phloroglucinol have been employed to characterize the profile and concentration of proanthocyanidins present in the extract under evaluation [63][64][65]. As previously reported, the proanthocyanidin molecules consisted of epicatechin units with mainly DP of 4 and 5 containing at least one A-type linkage [64].

Flow cytometry analysis of cellular apoptosis
NCI-H460 cells (1.5 × 10 6 cells) were incubated for 24 hours before PAC [0.25 or 50 µg/mL of media] or vehicle (<0.001% ETOH) was added. Apoptosis was evaluated in at least triplicate at each time point which included 2, 6, 12, 24, and 48 hours post-treatment using Annexin V-FITC staining methods and counted using a FACSCalibur flow cytometer with a minimum of 10,000 cells counted [66]. Analysis of apoptosis was performed using WinMDI software (Joseph Trotter; http://pingu.salk.edu.software.html) and ModFit LT software (Verity Software, Topsham, ME, USA).

Western blot analysis
NCI-H460 cells (1.0 × 10 6 ) were incubated for 24 hours, rinsed with PBS, treated with PAC (0.25 or 50 µg/mL) or vehicle (<0.001% ETOH) and harvested at 0, 3, 6, 24, 48 and 72 hours post-treatment. Cell lysates were prepared using Cell Signaling lysis buffer. Protein was quantified using the Quick Start Bradford Protein Assay Kit (BioRad) and equivalent protein amounts were loaded into precast NuPage Novex Bis-Tris 10% gels (Invitrogen). Immunoblot was performed using commercially available antibodies from Santa Cruz Biotechnology (Santa Clara, CA, USA) to proteins of interest including P21 (sc-6246), Cytochrome C (sc-13156), PARP1 (sc-8007), and GAPDH (sc-32233) as loading control. Expression values were determined by chemiluminescent immunodetection and normalized to GAPDH. Fold-change from baseline or first detection level was calculated. Positive fold-change values indicate increased expression.

Isolation of RNA and synthesis of cDNA
Total RNA was prepared from cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and quantity and quality were determined by Nanodrop using the Bioanalyzer 2100 capillary electrophoresis system (Agilent) at the Ohio State University (OSU) Microarray Core. Reverse transcription of purified RNA to cDNA was generated using 1.0 µg of total RNA, oligo(dT), and random hexamers (Applied Biosystems, Foster City, CA, USA) as primers for first strand cDNA synthesis under the following conditions: 30 °C for 5 minutes, 37ºC for 15 minutes, 42 °C for 60 minutes, 50 °C for 10 minutes, and 70 °C for 15 minutes. The first strand synthesis product was treated with RNase H (15 minutes, 37 °C) prior to use in the real-time reaction.

Microarray studies
NCI-H460 lung cancer cells (2.0 × 10 6 cells) were treated with vehicle or PAC [50 µg/mL] for 6 hours. One microgram of RNA per condition was reverse-transcribed, labeled by incorporating biotinylated nucleotides during in vitro transcription, and hybridized to the human U133 2.0 Plus chip per manufacturer recommendations (Affymetrix, Santa Clara, CA, USA). Specific transcripts bound to the corresponding oligonucleotide probes and the biotinylated cRNA bound fragments were detected using a streptavidin-antibody-phycoerythrin conjugate.

Validation by Real Time PCR
Real time PCR for differentially expressed genes of interest was performed utilizing RT² Profiler PCR Arrays (SABiosciences; Fredrick, MD, USA) and the iCycler IQ (Bio-Rad) to perform real-time PCR. Specifically, the apoptosis and cell cycle PCR pathway focused 'Profiler Arrays' were utilized permitting evaluation of 84 pathway linked probes for gene expression profiling. Relative changes in gene expression were calculated by 2 -∆∆Ct , where ∆∆Ct = ∆Ct (treated) -∆Ct (untreated). Data were normalized to expression levels of a combination of control genes including HPRT1, RPL13A and GAPDH.

Statistical and microarray analysis
Results are presented as the mean value ± SD for the apoptosis experiments. Data were evaluated for statistical significance using the Student's t test (two-sided, p < 0.05). Microarray data was analyzed utilizing previously described techniques [67][68][69][70][71]. Briefly, analysis was performed utilizing unbiased differential gene expression, comparing relative fluorescence intensities between arrays, and Affymetrix images were transformed into CEL files utilizing GCOS software (Affymetrix). Gene expression levels were estimated from GeneChip probe intensities using the WEDGE++ algorithm [69]. WEDGE++ computes p-values based on nonparametric probe-level multi-array chi-square tests for differential gene expression. Next, two-fold differentially expressed genes were analyzed using Expression Analysis Systemic Explorer (Ease software, DAVID 6.7, updated March 2009; http://david.abcc.ncifcrt.gov/) [67,71] to identify over-represented biologic themes classified by gene ontology categories. Signaling pathways were further investigated by viewing the PANTHER (Protein ANalysis THrough Evolutionary Relationships) biological processes which is linked to the DAVID website. PANTHER is a database of phylogenestic trees of protein-coding gene families supporting a number of database identifiers [71,72].

Conclusions
In summary, the present study demonstrates that cranberry proanthocyanidins significantly modulate cancer-related biological processes and key signaling pathways in NCI-H460 lung cancer cells following a single exposure at a behaviorally achievable concentration. PAC had highly significant and rapid apoptosis inducing effects and potent effects on multiple cell cycle linked genes resulting in decreased cell proliferation and increased cell death. Specifically, PAC increased P21 expression levels, which has been linked to apoptosis resistance; thus, PAC may hold promise as a chemopreventive agent during the early phases of carcinogenesis or may act to re-sensitize cancer cells to apoptosis and chemosensitivity. Further investigation of PACs potential to induce autophagy is also warranted given PACs rapid induction of cell death and up-regulation of p73, a gene linked to both type 1 and type 2 cell death. Mechanistic research on PACs cancer inhibitory potential is ongoing in a larger panel of aerodigestive tract cell lines.