Generation of Lung Adenocarcinoma DNA Aptamers for Cancer Studies

Lung cancer is the most lethal malignancy in the world, and each year thousands of people die from this disease. Early detection has proven to increase the 5-year survival for this cancer in general, independent of the origination site in the lung. To address this challenge, we have used cell-based SELEX (Systematic Evolution of Ligands by Exponential Enrichment) to select a panel of aptamers capable of distinguishing lung adenocarcinoma cells from normal lung epithelial cells. These aptamers bind at physiological and formalin-fixed conditions and display affinity for their targets with apparent Kd's in the nanomolar range. Our findings suggest that the selected aptamers have the potential to be used in clinical settings, as well as to improve classification of nonsurgical specimens, another current challenge in lung cancer.


Introduction
Lung cancer is the second most common type of cancer and the leading cause of cancer-related deaths worldwide [1]. In the United States, it is estimated that 226,160 new cases and 160,340 deaths will occur in 2012 (for Non-small cell and small cell combined). Compared to other cancer types, lung neoplasms are highly heterogeneous, with tumors displaying more than one subtype as a common feature [2]. The vast majority of lung neoplasms are carcinomas, which are generally classified as either non-small cell lung carcinomas (NSCLC) or small cell lung carcinomas (SCLC) on the basis of morphological analysis by stained histological samples [3][4]. NSCLC is the most common lung cancer type, comprising 85% of all lung cancer cases; yet it is a more passive cancer type. NSCLC is composed of three different subtypes: adenocarcinoma (ADC), squamous cell carcinoma (SCC), and large cell carcinoma (LCL). On the other hand, SCLC is less common, comprising 15% of all lung cancer cases, but it is more aggressive. Smoking is a risk factor heavily associated with lung cancer, specifically SCC [5][6][7][8][9][10]. Lung adenocarcinoma is commonly developed by patients who have never smoked, and genetic changes are often associated with its onset.
Since most people with lung cancer at the early stage do not display symptoms, more than 70% of lung cancer cases are diagnosed at later stages, for which the 5-year survival rate is small. Therefore, research aimed at early detection, which is critical to reducing mortality and morbidity, has turned to the development of suitable aptamers. Aptamers are short, single-stranded DNA or RNA oligonucleotides which are highly specific target recognition elements based on their unique three-dimensional shapes [11][12]. While the process known as SELEX (Systematic Evolution of Ligands by Exponential Enrichment) was originally used to select aptamers against targets such as purified proteins [13][14], cellbased SELEX has become the newest method of selecting aptamers against whole cells, especially those aptamers targeting surface proteins overexpressed in cancer cells.
Among their many advantages, aptamers have shown no, or extremely low, immunogenicity, permitting in vivo studies using these probes [15][16][17][18]. They have also been popularized as alternatives to antibodies, because of aptamers' low cost (no animals necessary for production), easy chemical modification, and cellular uptake capability. In addition, because aptamers are small in length with generally 15 to 100 nucleotides (nt), they have better tissue penetration compared to antibodies. In 2004, Macugen, an anti-VEGF (Vascular Endothelial Growth Factor) inhibitor, became the first aptamer approved by the Food and Drug Administration (FDA) for Age-Related Macular Degeneration (AMD) [19]. Other aptamers remain in clinical trials [20], and have demonstrated great potential in the biomedical field, including separation, drug delivery and target-probe measurement. This report describes the use of cell-SELEX to select a panel of aptamers capable of distinguishing between lung adenocarcinoma and normal lung epithelial cells.

Results and Discussion
Since their discovery, aptamers have been generated against different targets, including proteins, peptides, and living cells [21][22]. To isolate aptamers capable of differentiating lung adenocarcinoma cells from normal lung epithelial cells, we used the cellbased SELEX strategy. H23 lung adenocarcinoma and HBE 135-E6/E7 normal epithelial lung were used as positive and negative cell lines, respectively. An initial ssDNA random library containing approximately 10 14 different sequences of 80 nucleotides (nt) was enriched by sequential binding with the target cells, elution and subsequent amplification by PCR for 18 rounds. These DNA sequences could recognize H23 cell-surface membrane proteins which are potential markers for targeted therapy. In earlier rounds of the process, counter selection was introduced in order to remove possible sequences binding common proteins on both target and negative cell lines. This procedure was performed every other round throughout the selection. Sequences binding to target cells were eluted and PCR-amplified, after which ssDNA was recovered and used to monitor the selection process by flow cytometry. Because the ssDNA pools were enriched with sequences specific for the target, an increase in fluorescence intensity was first noticed in round 12 ( Figure 1A), indicating that those sequences showed better binding on the surface of H23 cells compared to the initial library. As the selection progressed, the fluorescence intensity of the subsequent pools gradually increased until a steady state in fluorescence intensity was observed in rounds 17 and 18 ( Figure 1B), indicating that maximum binding had been achieved. In contrast, such enrichment was not observed when the enriched pools were tested against normal lung epithelial cells ( Figure 1D), implying that sequences contained in those pools were capable of distinguishing lung adenocarcinoma cells from normal lung epithelial cells. Therefore, pools 14, 16 and 18 were selected to be sequenced as aptamer candidates ( Figure 1C).
The sequencing process began with the construction of the 454 sequencing library by PCR-addition of 454-specific primers to the 39-and 59-ends of the enriched pools. To identify the pool for each sequence, a unique identification code (middle identification code, MID) was also PCR-introduced to the 454 sequencing library. The insertion of primers in the enriched pools was confirmed by gel electrophoresis (data not shown). After 454 sequencing at the UF ICBR core, around 7,000 sequences were retrieved and analyzed. They were grouped on the basis of the MID corresponding to the same pool, and primers were removed by a Perl program. The sequences containing only the random region were then aligned using the online program MAFTT 6.0 [23], and six aptamer families corresponding to the most abundant sequences were chosen as aptamer candidates. They were chemically synthesized, biotin-labeled at the 39-end, purified by HPLC, and quantified. All aptamers displayed binding with H23 lung adenocarcinoma cells, while no significant binding was observed with the control cell line HBE 135 E6/E7, normal epithelial lung cells ( figure 2 and figure S1). These results indicate that successful negative selection had been carried out and that selected aptamers could distinguish between lung adenocarcinoma and normal lung epithelial cells, an outcome which confers these aptamers with the potential for use in lung cancer diagnosis, monitoring tumor progression and treatment, or other relevant applications.
All the selected aptamers showed binding affinities (apparent K d 's) to the H23 cell line in the nanomolar range (45-250 nM) ( Table 1 and figure S6). These results suggest that these aptamers will be widely applicable, as they tightly bind to their target. Further studies were carried out to characterize the selected aptamers. The binding was further tested against lung cancer cell lines, as well as other cell lines, including ovary and colon, as shown in Table 2. Aptamers EJ2, EJ4 and ADE1 displayed some recognition towards one or more of the following cell lines: TOV21G, DLD1, and H460. In the case of DLD1, a colon adenocarcinoma cell line, it was interesting to see some recognition by the selected aptamers, suggesting that a possible common cell-surface target is present in these two cell lines, since they belong to the same histological group. In a previous study in our lab, the target protein of Sgc8 aptamer was demonstrated to be PTK7 [24], a pseudo-kinase protein present in CEM cells (the target cell line for that selection), as well as other cancers, including ovary, lung, colon, breast and some leukemia cell lines, indicating that common proteins can be present as a result of cancer [25][26]. Aptamers, EJ4 and ADE1 also showed particular specificity towards the lung adenocarcinoma cell line with significant increase in fluorescence intensity with respect to a random sequence, but not to normal lung cells. These results suggest that these aptamers could be used for lung cancer studies.
Flexible binding of aptamers at different temperatures can expand their repertoire of applications. Since the selection was performed at 4uC, we performed binding assays at 25uC and 37uC ( Figure 3 and Figure S2). As shown in Figure 3 and Figure S3, aptamers ADE1, ADE2, EJ2, EJ5 and EJ7 conserved binding at 25uC and 37uC with fluorescence intensities similar to those at 4uC. This is particularly important in assays carried out under physiological conditions, such as those assessing in vivo applications.
Cancer detection relies on the examination of tissue from biopsies, which are fixed under chemical conditions to preserve the integrity and antigenicity of tumor samples for later use. An important assay in cancer diagnosis is immunostaining, in which freshly dissected tissues are fixed prior to treatment with probes, such as antibodies and aptamers, specifically for tumor markers. Previous studies have shown that aptamers are capable of labeling formalin-fixed paraffin tissue (FFPE) after deparaffinization and antigen retrieval [27][28]. To determine the binding capability of the selected aptamers under those conditions, cells were fixed with 10% formalin before incubation with labeled aptamers. As control for these experiments, two known aptamers for leukemia, Sgc8 and TD05, and their corresponding binding cell lines were used to show that the process of fixation does not produce any fluorescence signal. Therefore any fluorescence detected should be an indication of a binding event between the aptamers and its target. As shown in figure S3, both controls retained their initial binding profile, indicating that no artificial increment in fluorescence intensity occurred as a result of the fixing conditions. EJ2, EJ5, EJ7, ADE1, and ADE2 aptamers maintained similar binding to that displayed at 4uC (Figure 4) indicating that aptamers can bind to their targets even under fixed conditions. These results strongly suggest that selected aptamers have the potential to be used as recognition molecules in clinical samples.
During cell selection, it is expected that aptamers interact specifically with the surfaces of target cells; this behavior has been reported in several studies [29][30][31]. In our experiment, cells were treated with two enzymes, trypsin and proteinase K, for 3 and 10 minutes, washed with PBS, and then incubated with aptamers. All selected aptamers lost binding after treatment with both proteases ( Figure 5 and Figure S5. For all experiments, the fluorescence intensity was significantly reduced; however, the signal corresponding to assays with proteinase decreased to the background, indicating that the target protein was removed completely after treatment.

Conclusion
In conclusion, we have selected a panel of aptamers capable of distinguishing lung carcinoma from normal lung epithelial cells. These aptamers show high affinity towards the H23 cell line with apparent K d ' s in the nanomolar range, but no detectable affinity for normal lung epithelial cells. Aptamers also showed binding under physiological conditions, as well as after chemical fixation, suggesting that they can be applicable for in vivo experiments. Proteinase treatment indicated that all aptamers in this panel bind to proteins on the target cell surface. All these results suggest broad potential applications of selected aptamers due to their specificity for cancerous tissues but not for healthy cells. These aptamers can also be used as molecular probes in clinical samples, such as freshly extracted tumors and preserved histology specimens.

Library Design
The primers were design to satisfy the following characteristics: a minimum hairpin structure, similar melting temperature (T m ) and minimal base pairing. The primers and an 80-mer library were designed using the IDT Oligo Analyzer 3.1 software [32]. The forward primer was labeled with Fluorescein Isothiocyanate (FITC) at the 59-end, and the reverse primer was labeled with Biotin at the 59-end. The library consisted of a randomized 44-nt region flanked on the 59-end by the FITC-labeled primer and flanked on the 39-end by the complementary unlabeled strand of the reverse primer.

Instrumentation and Reagents
Libraries and primers were synthesized using the 3400 DNA synthesizer (Applied Biosystems). All reagents for DNA synthesis were purchased from Glen Research. DNA sequences were purified by reversed phase HPLC (Varian Prostar using a C18 column and acetonitrile/triethylammonium acetate as the mobile phase). PCR was performed on a Biorad Thermocycler, and all reagents were purchased from Takara. The monitoring of the selection process, binding assays, and determination of the dissociation constants for the selected aptamers were performed by flow cytometric analysis using a FACScan cytometer (BD Immunocytometry Systems).

Cell Culture and Buffers
A total of eight established cell lines were used in this project, all purchased from the American Tissue Culture Collection (ATCC). H23 (CRL-5800) adenocarcinoma NSCLC was chosen as the positive cell line, while the negative cell line chosen was HBE135-E6/E7 (CRL-2741), normal human bronchial epithelial cells. The H23 cell line was maintained in RPMI-1640 (ATCC) culture medium supplemented with 10% Fetal Bovine Serum (FBS heatinactivated) and 1% penicillin-streptomycin. The normal bronchial lung cell line was maintained in a Keratin Serum-Free Medium supplemented with 5 ng/mL human recombinant EGF, 0.05 mg/mL bovine pituitary extract (Invitrogen), 0.005 mg/mL insulin, and 500 ng/mL hydrocortisone (Sigma-Aldrich). Cells were incubated at 37uC under 5% CO 2 atmosphere. Other cell lines used for the selectivity assay were the CAOV3 and TOV21G ovarian cancer cell lines, A549 lung adenocarcinoma, H520 lung squamous cell carcinoma, H460 large cell lung carcinoma, and DLD1 colon adenocarcinoma, and all were maintained according to ATCC specifications. During selection, two buffers were used: Washing Buffer (WB) (glucose 0.45% w/v and MgCl 2 5 mM in PBS) and Binding Buffer (BB) (1 mg/mL tRNA and 1 mg/mL

In vitro Selection
Because both cell lines used during the selection, adenocarcinoma and normal human bronchial epithelial cells, are adherent cell lines, the selection was performed on cell monolayers. About 20 nmol of the synthesized library was dissolved in 700 mL of binding buffer. Before the process was initiated, the DNA pool was denatured by heating at 95uC for five minutes, followed by rapid cooling on ice. This forced the DNA sequences to adopt the most favorable secondary structures. The DNA library was then incubated with approximately 3610 6 target cells (H23) at 4uC for 30 minutes. Subsequently, the cells were washed 3 times with washing buffer to remove unbound sequences. Afterwards, the bound sequences were recovered by heating at 95uC for 10 minutes and then centrifuging at 14,000 rpm to remove cell debris.
The supernatant containing the DNA sequences was collected, and the selected pool was PCR-amplified using FITC-and biotin-labeled primers. Then, the generation of single strand DNA (ssDNA) was achieved by incubation with streptavidin-coated sepharose beads, obtaining the biotinylated strand. Counter selection was carried out after observing some enrichment with the target cells. In order to select aptamers with high specificity and selectivity, the stringency of the washes (number of washes, wash time, and volume of washing buffer) was increased in subsequent rounds. The enrichment of the pools was monitored using flow cytometry, sequenced with 454 technology, and analyzed for aptamer candidates.

Flow Cytometric Analysis
Flow cytometry was used to monitor the enrichment of ssDNAbound sequences within the pools during the selection process, as well as to evaluate the binding affinity and specificity of the selected aptamers. The cultured cells were washed with WB before and after incubation with the FITC ssDNA pool or selected DNA sequences. Fluorescence intensity was determined on a FACScan cytometer (BD Immunocytometry).

Sequencing and Analysis
After 18 rounds of selection, enriched pools 14, 16, and 18 were chosen for sequencing. 454-specific primers and MID were PCRamplified and added to each sequence contained in each pool, yielding a 125-bp product. The reactions were confirmed by gel electrophoresis, cleaned using a PCR purification kit (Qiagen), and submitted to the ICBR core at the University of Florida for analysis.
Around seven thousand sequences were retrieved and analyzed. First, sequences were grouped on the basis of their MID to determine the corresponding enriched pool. Second, primers and MID were removed by the Perl program in order to leave only the random portion of the sequence for homology analysis with the MAFFT 6.0 program.

Binding Assays
Different families of sequences retrieved from multiple sequence analyses were synthesized, biotin-labeled at the 39-end, and tested for binding with the positive and negative cell line. In addition, the binding selectivity of each candidate was determined by incubating a 250 nM aptamer solution with 4610 5 target or counter-cells for 30 minutes at 4uC. Cells were washed twice with WB and incubated with streptavidin PE beads for 20 minutes at 4uC. After washing, the cells were suspended in 200 mL WB. The fluorescence was determined with a FACScan cytometer by counting 3610 4 events. A randomized 80-mer sequence was used as a control.

Binding Assays with fixed cells
To determine the ability of the aptamers to bind to fixed cells, procedures similar to those described above were followed, with the exception of the initial treatment of the cells. Briefly, adherent H23 cells were detached from the dish by incubation with nonenzymatic cell dissociation solution, and then fixed with 10% formalin solution for 15 min at 4uC, washed, and suspended in binding buffer.

Selectivity and Specificity Assays
To determine the specificity of these aptamers, different cell lines, including CAOV3, DLD1, A549, H520, H460, and TOV21G, were used in binding assays. Fluorescence intensity was measured in order to confirm binding. All experiments followed the flow cytometry experimental procedure described above.

Temperature effect on aptamer binding
Selection was performed at 4uC.To determine if temperature would affect the binding between aptamer and the target cells, binding assays were carried out at two additional temperatures, 25uC and 37uC.

Determination of the dissociation constant
The affinity of the aptamer and its target was evaluated by saturation assay. A sample containing 5610 5 H23 cells was washed and incubated with different concentrations of aptamer until saturation was achieved. All binding assays were repeated three times, and the mean fluorescence intensity was calculated by subtracting the fluorescence intensity of a scrambled sequence. Data were collected, and the dissociation constant (K d ) was obtained by fitting to a single binding site saturation model using SigmaPlot 11.2v (Jandel, San Rafael, CA).

Protease Digestion
Trypsin and Proteinase K treatment. H23 cells were washed twice with PBS and then incubated with 3 mL of either 0.05% trypsin/0.53 mM EDTA in Hank's balanced salt solution (HBSS), Cellgro, or 0.1 mg/mL proteinase K in PBS at 37uC for 3 and 10 minutes. FBS was added to quench the proteinase activity. Cells were washed with WB and subsequently used for the binding assays described above. Figure S1 Characterization of selected aptamers. Flow cytometry assay for the binding of the aptamers ADE1, ADE2, EJ2, EJ5 and EJ7 with H23 (target cell line) and HBE135 E6/E7 (negative cell line). The green curve represents the background binding of a random sequence (library). (TIF) Figure S2 Aptamer binding in physiological conditions. Flow cytometry assay for the binding of aptamers ADE1, ADE2, EJ4, EJ5 and EJ7 with H23 (target cell line) at 25uC and 37uC. In  Figure S4 Binding assays after fixation with 10% formalin. Binding assay of aptamers with H23 (target) cells pre-fixed with 10% formalin. Left columns show the binding of aptamer ADE1, ADE2, EJ2, EJ5 and EJ7 with untreated cells at 4uC. Right column shows the binding of the same aptamers EJ5 with fixed cells. The light green curve represents the background binding of a random sequence (library). (TIF) Figure S5 Binding Assays after proteinase treatment. Flow cytometry assay for aptamers ADE1 and ADE2, EJ5 and EJ7 after treatment with proteases; untreated cells were used as positive control. (A) and (C) Cells treated with trypsin for 3 and 10 min prior binding with aptamers ADE1 and ADE2, EJ5, and EJ7

Supporting Information
respectively. (C) and (D) Cells treated with proteinase K for 3 and 10 min prior binding with aptamers respectively. (TIF) Figure S6 Apparent Kd's for selected aptamers. Saturation binding curves for selected aptamers ADE1, ADE2, EJ2, EJ5 and EJ7. Cells were incubated with different concentrations of the aptamer in triplicate. The mean fluorescence intensity of unselected library was subtracted from each corresponding aptamer concentration. (TIF)