Overexpression of MicroRNA miR-30a or miR-191 in A549 Lung Cancer or BEAS-2B Normal Lung Cell Lines Does Not Alter Phenotype

Background MicroRNAs (miRNAs) are small, noncoding RNAs (ribonucleic acids) that regulate translation. Several miRNAs have been shown to be altered in whole cancer tissue compared to normal tissue when quantified by microarray. Based on previous such evidence of differential expression, we chose to study the functional significance of miRNAs miR-30a and -191 alterations in human lung cancer. Methodology/Principal Findings The functional significance of miRNAs miR-30a and -191 was studied by creating stable transfectants of the lung adenocarcinoma cell line A549 and the immortalized bronchial epithelial cell line BEAS-2B with modest overexpression of miR-30a or -191 using a lentiviral system. When compared to the corresponding controls, both cell lines overexpressing miR-30a or -191 do not demonstrate any significant changes in cell cycle distribution, cell proliferation, adherent colony formation, soft agar colony formation, xenograft formation in a subcutaneous SCID mouse model, and drug sensitivity to doxorubicin and cisplatin. There is a modest increase in cell migration in cell lines overexpressing miR-30a compared to their controls. Conclusions/Significance Overexpression of miR-30a or -191 does not lead to an alteration in cell cycle, proliferation, xenograft formation, and chemosensitivity of A549 and BEAS-2B cell lines. Using microarray data from whole tumors to select specific miRNAs for functional study may be a suboptimal strategy.


Introduction
The significance of non-coding RNAs in the regulation of cellular processes has come to be increasingly appreciated [1]. The best studied non-coding RNAs are microRNAs (miRNAs), 18-25 nt-long small RNA that epigenetically regulate translation by binding to a complementary ''seed'' sequence common to their ''target'' mRNA [2]. As this complementarity does not have to be perfect, a single miRNA can regulate the expression of several hundred genes simultaneously [3]. The explosion of microarray technology has led to its application in miRNA genomics. Therefore, before the function of a significant proportion of miRNAs were known, miRNA profiling of several normal and abnormal human tissues have been performed [4]. The distinct difference between the data obtained between miRNA profiling and mRNA profiling is in the magnitude of differential expression seen in the comparison groups. Whereas several fold changes are commonly seen in mRNA expression using these technologies, the fold changes with miRNA experiments are more modest and are typically in the 1-to 2-fold range. Therefore, experiments to study the biological relevance of these modest fold changes are important to evaluate the significance of these changes. Based on previously published miRNA microarray data demonstrating differences in expression in human cancers compared to the corresponding normal tissue, two miRNAs were chosen in this study for such biological experiments. These two miRNAs were chosen because of consistent alterations in their expression between normal and cancerous tissue of various organs, a bioinformatic analysis of their predicted targets that suggested important roles in cell proliferation and migration and little published work on their functional roles in lung cancer.
MiR-30a is important for the development of the prefrontal cortex via the regulation of brain-derived neurotrophic factor (BDNF) [12] and in vertebrate hepatobiliary development [13]. Other experimentally confirmed targets of miR-30a are adenylate kinase (Ak1) and the GW182 protein (Tnrc6a), a component of the RNA interference silencing complex [13].
MiR-191 is located on human chromosome 3p21.31 and is generated from an intronic transcriptional unit that can produce two miRNAs: miR-191 and -425 [5]. Only a single mature form of miR-191 is known to exist. MiR-191 expression is up-regulated in pancreatic [14], colon, lung and prostate [7] cancers. It is also upregulated in AML patients with abnormal karyotypes [11], and down-regulated in patients with CLL [6]. MiR-191 expression is also altered in lungs exposed to cigarette smoke [15]. No mRNA targets of miR-191 have been experimentally verified.
In this study, we evaluate the phenotypic changes seen by constitutive over-expression of miR-30a and -191, on a lung adenocarcinoma cell line (A549) [16] and an immortalized bronchial epithelial cell line (BEAS-2B) [17]. The A549 cell line was chosen because of the existence of a large amount of experimental data on its use in various assays. The BEAS-2B cell line was chosen because it is the immortalized cell line closest to normal bronchial epithelium. As the effects of overexpression of a microRNA is context-specific, we compared the alterations in phenotype of malignant and non-malignant cells.

Ethics Statement
All in vivo studies were approved by the Institutional Animal Care and Use Committee of the Roswell Park Cancer Institute.

Generation of Stably Transfected Cell Lines
Single-stranded DNA oligonucleotides with human pre-miR-30a or -191 (miRBase accession IDs MI0000088 and MI0000465, respectively) sequences and with restriction enzyme site overhangs were obtained from Integrated DNA TechnologiesH (Coralville, IA). Complementary sequences were annealed and the resulting double-stranded DNA was ligated to Xho I/Not I-digested pLemiR TM vector (Open BiosystemsH, Huntsville, AL). A549 cells were then infected with plasmids using the Trans-Lentiviral TM GIPZ packing system (Open Biosystems; Huntsville, AL) according to the manufacturer's protocol. Briefly, TLA-HEK293TTM cells were transfected using Arrest-In TM with 37.5 mg plasmid DNA in serum free medium for 4 hours. Media was then replaced with serum containing media for 36 hours. Media were collected, centrifuged to remove cell debris and used to transfect A549 and BEAS-2B cells. Forty eight hours after addition of virus, transfected cells were selected for by adding 2 mg/ml puromycin to growth medium.

Real Time RT-PCR (qPCR) for Quantification of Small RNAs
Total RNA (20 ng), isolated from cells using PureLink TM Micro-to-Midi total RNA isolation kit (Invitrogen) according to the manufacturer's protocol, was reverse transcribed using TaqMan TM reverse transcription kit (Applied BiosystemsH, Foster City, CA) and RNA-specific primers provided with TaqMan TM microRNA assays (Applied BiosystemsH) in 15 ml, with annealing

Cell Proliferation Assay
Cell proliferation assays were performed using the Cell Titer 96H kit (Promega; Madison, WI) which uses the conversion of MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt) to formazan to measure cell viability. Briefly, cells were plated on 96 well plates at 4,000 cells in 100 mL per well. Cells were harvested in quadruplicate at various time points for 150 hours. To the harvested cells, 20 mL of assay reagent was added and viable cells were measured by absorbance at 450 nm 1 hour later on a Multiskan Plus (MTX Lab Systems; Vienna, VA) plate reader. Absorbance values were normalized to media control.

Adherent Colony Formation
Cells in culture were trypsinized and plated in duplicate in 6well plates at 200 and cells per well for A549 and 500 cells per well for BEAS-2B cell line derivatives respectively. Cells were allowed to grow at 37uC, 5% CO2 with media changes every 3-4 days until colonies were visible by eye. Media was aspirated and cells were stained with 0.2% methylene blue in 40% methanol for 30 minutes. Cells were washed and plates were scanned using an Epson Perfection V700 Photo scanner as a transparency. Quantitation of colonies was done using ImageJ software (Research Services Branch, National Institute of Mental Health, Bethesda).

Soft Agar Colony Formation Assay
Base agar/media mix was added to 6-well plates (final concentration of 0.7% agar, 1x +DMEM, 10% FBS, 2 mg/ml puromycin) to coat the plate and allowed to polymerize. Cells suspensions were made to 100 cells/ml in media/agar mix (final concentration of 0.35% agar, 1x +DMEM, 10% FBS, 2 mg/ml puromycin). 1.5 ml of the cell suspension was plated on top of the base agar in each well and allowed to polymerize. 2 ml complete media was added on top of the agar once polymerized. The plates were put in 37uC, 5% CO2, and overlay media was changed every 3-4 days. Cultures were grown until a few colonies were visible by eye, 2 mg/ml Thiazolyl Blue Tetrazolium Bromide in PBS was added to the cells and were returned to the incubator until colonies were stained. Once colonies were stained blue, the stain was removed from the wells and pictures were taken using Cannon Powershot A590IS camera.

In Vitro Scratch Assay
To estimate differences in cell migration, an in-vitro scratch assay was performed. Cross-hairs were drawn on the bottom of 6well plates for orientation for imaging. Cells were plated in triplicate wells of at high density to reach 100% confluence 24 hours later. A scratch was then made directly next to the vertical line drawn on the plate using moderate and consistent pressure using a 10 mL tip and the plate lid as a straight edge. Scratches were imaged immediately on an Axio Observer microscope (Zeiss; Maple Grove, MN) at 5X magnification and orientation to the cross-hairs noted. Cells were then returned to 37uC, 5% CO2 and imaged at various time points until all scratches were closed.

Xenograft Formation Assay
All in vivo studies were approved by the Institutional Animal Care and Use Committee of the Roswell Park Cancer Institute. Control and over-expressing A549 cell lines (6.06105 cells) were injected subcutaneously between the shoulder blades of 4-5 week old SCID mice (5 mice in each cohort). Tumors were monitored 2-3 times per week by caliper measurements until the largest tumor was 2 cm in longest direction, at which point all mice were sacrificed. Tumors were dissected and weighed.

Cell Cycle Analysis
Cells grown to 70%-90% confluence were detached by trypsinization, fixed in 70% ethanol at 4u C for 1-2 days, washed with PBS, and incubated at a density of 122610 6 cells/ml with 0.3 mM 4,6-diamidino-2-phenylindole dihydrochloride (DAPI; MP BiochemicalsH, Solon, OH) in PBS at room temperature in the dark for 100 min. After washing once with PBS, DAPI fluorescence from cells was assayed by flow cytometry using an LSR II (BD BiosciencesH, San Jose, CA) instrument equipped with a 408 nm violet laser diode and a 450/50 nm emission filter.

Drug Sensitivity Assay
After cells were grown to 50%-60% confluence in 96-well plates (5 wells per cell-line), the medium was replaced with that containing 1 mg/ml doxorubicin hydrochloride (Enzo Life Scien-cesH, Farmingdale, NY) or cisplatin (CalbiochemH, San Diego, CA). After 1.5-2 days of culture, viability of cells was measured as described for the cell proliferation assay and compared to those that were not exposed to the drugs.

Creation of Stable Cell Lines with Over-Expression of miR-30a and -191
Stable cell lines overexpressing miR-30a and -191 were generated using a lentiviral system as described above. MiR-30a was overexpressed ,2 fold and ,3 fold in both A549 (A-30a) and BEAS-2B (B-30a) respectively compared to controls (A0 and B0 respectively). Compared to controls, miR-191 was overexpressed ,2 fold and ,7 fold in A549 (A-191) and BEAS-2B (B-191) respectively ( Figure 1). Overexpression was also confirmed by visualization of RFP containing cells by microscopy.

Pathway Enrichment Analysis
Target prediction by the miRanda algorithm identified 731 targets for miR-30a-5p and 570 targets for miR-191. The top ten pathways identified by the pathway enrichment analysis included pathways influencing cell cycle, cell proliferation, drug resistance and cell motility, among others (Table 1).

Over-Expression of miR-30a or -191 Does Not Alter Measures of Proliferation Both In Vitro and In Vivo in A-549 and BEAS-2B Cell Lines
Pathway enrichment analysis suggested that miR-30a and -191 influenced cell cycle and cell proliferation. To examine the effect Figure 3. Analysis of migration and cell cycle with overexpression of miR-30a and -191. A. Cell migration of A-30a and B-30a is increased compared to controls; no effect of overexpression of miR-191 is seen. B. Cell cycle analysis of stable transfectants demonstrating no difference between cells over-expressing miR-30a and -191 compared to controls. doi:10.1371/journal.pone.0009219.g003 of overexpression of miR-30a and -191 on A549 and BEAS-2B, growth curves were plotted based on measurements of cell viability at different time points over a 150-hour period. No difference in growth curves were seen in A-30a and A-191 compared to A0 and in B-30a and B-191 compared to B0 (Figure 2A, 2B). Cell proliferation in different cell lines was also compared using adherent colony formation. No difference was seen in either number or size of adherent colonies in cell lines over-expressing miR-30a or -191 compared to the corresponding control cell line ( Figure 2C).
Anchorage independence is one of the hallmarks of transformation. We sought to determine the effect of overexpression of miR-30a or -191 on anchorage independent growth of A549 and BEAS-2B cells using the soft agar colony assay as described above. All lines formed very few colonies that were visible by microscopy with only 2-3 colonies visible by naked eye. No differences were seen after staining or by microscopy (data not shown).
Xenograft formation is thought to represent a biologically relevant measure of tumor aggressiveness. We assessed the effect of overexpression of miR-30a or -191 on xenograft formation in SCID mice by A549 cells. When compared to the control cell line (A0), A-30a and A-191 do not form larger xenografts ( Figure 2D). The increase of tumor volume over time is also no different in these cell lines compared to control (data not shown). As the BEAS-2B cell line is non-tumorigenic, we did not assess xenograft formation using B0, B-30a or B-191.
The Effect of Over-Expression of miR-30a or -191 on Cell

Migration and Cell Cycle Distribution of A-549 and BEAS-2B Cell Lines
Increased cell migration is an important mechanism that is thought to increase metastatic potential of cancer cells. This may be independent of cell proliferation rates. Therefore, we studied the effect of miR-30a or -191 on cell migration of A549 and BEAS-2B cell lines using the scratch assay as described above. There is a modest increase in the migration of A-30a and B-30a compared to A0 and B0 respectively. There was no significant difference in the migration distance between cell lines overexpressing miR-191 when compared to the corresponding control cell lines ( Figure 3A).
An examination of the predicted targets of miR-30a and -191 and the cellular processes influenced by them shows a statistically significant enrichment of pathways involving the cell cycle. Therefore, we studied the effect of overexpression of miR-30a or -191 on cell cycle distribution. Flow cytometry after staining with DAPI was used to perform cell cycle analysis of all cell lines. Compared to the controls, cell lines over-expressing miR-30a or -191 do not show significant changes in the cell cycle distribution ( Figure 3B).
Over-Expression of miR-30a or -191 Does Not Alter Chemosensitivity to Doxorubicin or Cisplatin Doxorubicin and cisplatin are commonly used chemotherapeutic agents. A549 is moderately sensitive to cisplatin and PLoS ONE | www.plosone.org doxorubicin (http://dtp.nci.nih.gov), enabling the assessment of both increased and decreased sensitivity to these drugs. Overexpression of miR-30a or -191 does not alter the chemosensitivity of A549 and BEAS-2B to either drug ( Figure 4).

Discussion
MiRNA changes are commonly seen in cancerous tissue compared to their normal counterparts. The magnitude of these changes are modest and their functional significance in unknown. One of the difficulties in the study of miRNAs are that one miRNA has several hundred predicted targets and these bioinformatic predictions are not very accurate. Also, the role of miRNAs in a cell depends on the genetic constitution and the transcriptome of a cell at any given time and this complexity is not captured simply by their forced over-expression. However, phenotypic alterations secondary to forced over-expression can provide clues to understand their contribution to the malignant phenotype. In this study, we generated stable transfectants that constitutively over-express miR-30a or -191 and do not demonstrate significant changes in phenotype in the various assays that were performed. A number of possible explanations can explain this lack of alteration of phenotype. First, it is possible that the level of expression of the miRNAs achieved is not enough to create a difference. Even though the fold changes seen by microarray experiments is typically modest, such experiments may underestimate actual changes in malignant epithelial cells due to the dilution of changes in malignant epithelial cells by stromal cells that constitute a significant proportion of pathological specimens. Also, the magnitude of changes seen by microarrays is often smaller than the corresponding changes seen by RT-PCR. Therefore, the actual changes in malignant cells may be more than that achieved in our model system. Conversely, it is possible that the changes seen in miRNA microarrays spuriously overestimate changes. For example, in a study by Peltier et al [18] that carefully evaluated several candidate miRNAs for stability of expression across five types of cancer and normal tissues by RT-PCR, miR-191 was so stably expressed that it was proposed as a gene to be used for normalization. This argues against a role for miR-191 in carcinogenesis. Therefore, selecting miRNAs based on microarray data may not be a good strategy. Third, all miRNAs may not be able to result in phenotypic changes on their own; their effect may be context specific and may need to be coupled with alterations of expression of either other miRNAs or mRNAs to alter phenotype.
In summary, we conclude that over-expression of miR-30a or -191 does not lead to an alteration in cell cycle, proliferation, xenograft formation and chemosensitivity of A549 and BEAS-2B cell lines. Using microarray data from whole tumors to select specific miRNAs for functional study may be a sub-optimal strategy. It is important to confirm the functional significance of changes seen in miRNA expression data in specific tumor types to define the biological roles of specific miRNAs.