4.1. Whole-Cell SELEX Strategy
The use of complex target for aptamer selection was introduced for the first time by Morris et al. [9
] in 1998. They used human red blood cell membrane preparations (RBC ghosts) instead of purified membrane proteins to isolate multiple ssDNA molecules with high affinity for different cell membrane targets. This work opened the path to the development of whole, living cell SELEX protocols. Generally, these protocols have been adopted to isolate aptamers able to recognize a known target of interest or a specific cell phenotype (Figure 2
a,b). In both cases, a fundamental aspect is the inclusion of a counter-selection step to avoid the parallel enrichment of aptamers for unwanted targets. The negative selection step is introduced before the positive selection at each round, allowing filtering out sequences against those molecules commonly expressed on both the target and control cell lines.
Many cell-SELEX strategies have been used to isolate aptamers against a known protein of interest by using in the positive selection step a cell line over-expressing the target (Figure 2
a). For example, aptamers that specifically recognize the transforming growth factor-β type III receptor (TGFβRIII) have been generated by such an approach [10
]. The authors used Chinese hamster ovary (CHO) cells ectopically expressing TGFβRIII as target and isolated a RNA aptamer with a very good dissociation constant (1 nM range), able to efficiently recognize the cell surface receptor. The aptamer is also able to inhibit the interaction between TGFβRII and TGF-β2 ligand in vitro, showing potential applicability as a tool to investigate the receptor function and as therapeutic agent.
This strategy employs mock cells for counter-selection. A novel selection protocol, referred as “Isogenic cell-SELEX” (Icell SELEX), has been recently proposed for the isolation of aptamers against a known target [11
]. The authors used HEK293 cells in which the integrin alpha V (ITGAV), a transmembrane receptor expressed in almost all cell lines, is depleted (in the counter-selection step) or over-expressed (in the positive selection step). By such an approach, they easily isolated several anti-ITGAV aptamers.
A variant of the cell-SELEX approach, called hybrid- or crossover-SELEX, has been also proposed in order to enhance the success of the screening and the aptamer specificity for the target, by switching during the selection from cell-SELEX to purified protein SELEX. Such an approach revealed to be effective in isolating two high-affinity Tenascin-C (TNC) aptamers [12
]. The adapted protocol included a first enrichment of the pool on U251 glioblastoma cells over-expressing TNC, followed by a further enrichment against the recombinant purified protein. An inverted two-stage selection scheme, combining a first selection against the recombinant protein and a subsequent enrichment by cell-SELEX, has been as well developed to isolate anti-transferrin receptor (TfR) aptamers [13
]. The minimized anti-TfR aptamers were used to functionalize stable nucleic acid lipid particles (SNALPs), containing targeted siRNA achieving enhanced and specific uptake and silencing. A dual SELEX protocol was also applied to isolate DNA aptamers targeting the G-protein-coupled cholecystokinin B receptor (CCKBR) that is constitutively expressed by pancreatic ductal adenocarcinomas (PDACs) [14
]. The adapted selection strategy combines cycles against “exposed” CCKBR peptides and CCKBR-expressing PDAC cells. Among the isolated sequences, one aptamer (termed AP1153) shows a Kd of 15 pM and is internalized into target cells in a receptor-mediated manner, efficiently acting as carriers of fluorescent nanoparticles (NPs) to PDAC tumors in vivo. Similarly, Martínez Soldevilla et al. [15
] used a hybrid SELEX to isolate aptamers targeting the Multidrug Resistant-associated Protein 1 (MRP1), a 17 trans-membrane protein that has been correlated with resistance to chemotherapy in several cancers. The developed protocol includes 10 rounds of selection against a MRP1-peptide and a last round of cell-based selection using the chemotherapy-resistant H69AR cell line, showing a high MRP1 expression, as target. The selected aptamer was then used to generate a MRP1-CD28 bivalent aptamer to deliver the CD28 costimulatory signal of tumor-infiltrating lymphocytes to MRP1-expressing tumors. The bi-specific aptamer enhances costimulation in chemotherapy-resistant tumors and reduces tumor growth in melanoma-bearing mice upon systemic administration. An interesting combinatory screening in vitro and in vivo has been more recently proposed by Zhu et al. [16
] to isolate human epidermal growth factor receptor 2 (HER2)-targeting DNA aptamers for preclinical HER2 imaging in ovarian cancer. The authors first performed an in vitro protein-based SELEX on HER2 purified extracellular domain, and then used the resulting DNA pool for a cell-based selection on HER2- over-expressing SKOV3 cells. The best candidates have been radiolabeled with 18
F and screened in vivo in mice bearing SKOV3-tumors by PET imaging. This strategy led to isolate two aptamers enabling a rapid visualization of HER2-positive tumors with a good tumor uptake and tumor-to-muscle ratios.
An alternative cell-SELEX strategy (referred as “differential cell- SELEX”) (Figure 2
b) has been developed to isolate aptamers able to recognize a specific cell phenotype, rather than a single specific target of interest. This strategy offers the possibility to select multiple ligands discriminating between even closely related cell types, without any prior knowledge of the target. Briefly, the procedure consists of the incubation of the starting library on a non-target cell line (with undesired phenotype, negative selection step) followed by the recovery of unbound oligonucleotides that are, then, incubated on cells with the desired phenotype (positive selection step). By using such an approach, ssDNA aptamers able to specifically distinguish T-cell acute lymphocytic leukemia (ALL) cells from B-cell lymphoma cells [17
] or small lung cancer cells versus large cell lung cancer [18
] have been selected. A similar strategy has been used to isolate aptamer-based probes selective for colorectal cancer cells [19
] or nasopharyngeal carcinoma (NPC) [20
]. In the last study, the authors performed a cell-SELEX on NPC 5-8F (target cells) using nonmalignant human nasopharyngeal (NP) epithelial NP69 cells (non-target cells) in the counter-selection step. They isolated four aptamers able to discriminate between NPC and NP cells and then employed an aptamer-based affinity purification combined with mass spectrometry, identifying CD109 as the target of the S3 aptamer.
More recently, Yoon et al. [21
] applied a “blind” SELEX protocol to pancreatic cancer using Huh7 (hepatocarcinoma cells) as negative cells and PANC-1 cells as positive selection. By chromatography tandem mass spectrometry, the authors demonstrated that one of the aptamers (P15) binds the intermediate filament vimentin, a biomarker of epithelial–mesenchymal transition, and significantly inhibits cell invasion. In addition, the differential cell-SELEX strategy has been applied to discriminate between cell types with different properties (i.e., tumorigenesis and metastatic potential). In this context, our laboratory reported the isolation of a panel of aptamers able to bind human malignant glioblastoma (GBM) cells, discriminating them from non-tumorigenic GBM cells [22
]. More recently, Li et al. proposed metastatic-cell-based SELEX to isolate aptamers specific for metastatic cancer cells. The authors applied this strategy to colon cancer, using as target of the selection SW620 cells, derived from metastatic site lymph node. By such an approach, they identified an aptamer (XL-33-1) with high affinity and selectivity for target cells. [23
]. Most importantly, tissue imaging experiments with FAM-labeled XL-33-1 revealed its ability to specifically identify the metastatic tumor or lymph node tissues, thus showing its great potential as a molecular imaging agent for early detection of colon cancer metastasis. Similarly, Yuan et al. [24
] used metastatic colorectal carcinoma cell line LoVo as selection target and non-metastatic colorectal carcinoma cell line SW480 for the counter-selection, identifying a new aptamer (J3) able to recognize a metastasis-related membrane protein. This aptamer was labeled with Cy5 and used as effective imaging contrast for colorectal carcinoma metastasis with a detection rate of 73.9%. Aptamers able to discriminate high metastatic versus low/non metastatic cells derived from prostate cancer or hepatocellular carcinoma, have been also recently selected by cell-SELEX [25
]. Selected aptamers were labeled and successfully used for metastatic tissue imaging. Then, Kim et al. [27
] proposed an interesting differential protocol, performing repeated cycles of positive selection on tumor-initiating cells (TICs) and negative selection for non-TICs and human neural progenitor cells. This strategy led to isolation of a pool of sequences able to specifically bind to and internalize into TICs, thus showing a potential applicability for TIC targeting and imaging. Further, a DNA aptamer (MS03) specifically recognizing mammospheres was isolated following 13 rounds of selection on MCF-7-derived mammospheres. In the proposed protocol, normal breast epithelial MCF-10A and Salinomycin-treated MCF-7cells were used in the counter-selection step [28
]. The authors demonstrated that MS03 aptamer was effective in isolating breast CSCs by flow cytometry. More recently, a modified cell-SELEX method has been applied also to stem-enriched cancer cells in pancreatic cancer by performing positive selection on spheres derived from pancreatic cancer cell line, HPAC, and negative selection on pancreatic normal cell line, HPDE. The approach led to the isolation of two sequences with high specificity and good affinity for the CSC population [29
]. Notably, one of the sequences (labeled with Cy5) allows detection of CSC marker over-expressing circulating tumor cells, isolated from blood samples of metastatic pancreatic cancer patients.
In order to investigate cancer cell drug resistance, Zhang et al. [30
] applied a differential SELEX protocol using doxorubicin-resistant MCF-7R breast cancer cell line as target of the selection and its parental cell line (MCF-7) for counter-selection. Interestingly, by such an approach the authors isolated an aptamer (M17A2) able to recognize intercellular connections (tunneling nanotubes), thus representing a novel probe for cell-cell communication studying.
All these differential cell-SELEX approaches have the potential to provide multiple targeting ligands with a great applicability as specific probes and tools for biomarker discovery. Nevertheless, a limit arises from the absence of notion of the target recognized by the selected aptamers. Therefore, subsequent target identification strategies based on proteomics analysis [31
], which are often very complicated and time-consuming, are required.
4.3. 3D Cell-SELEX and In Vivo SELEX
One of the drawbacks of cell-based selection strategies is that it may happen that in vitro selected aptamers fail to be effective in vivo. Indeed, it is possible that protein target conformation is affected by the target’s environment that can change in vivo. In order to mimic the microenvironment in vitro, recently Souza et al. [69
] developed a novel cell-based strategy by using spheroid cells of human prostate cancer in 3D cell culture as target of the selection. The authors performed a first round of negative selection against the non-tumor cell line RWPE-1, followed by eight rounds of selection against PC-3 cell line. By such an approach, they generated an aptamer able to specifically bind prostate tumor cells, with a dissociation constant in the nanomolar scale and potential application in screening assays for prostate cancer.
Most importantly, even very complex targets, including tumors implanted in mice (in vivo-SELEX), have been used to select aptamers. For example, Mi et al. [70
] screened a nuclease-resistant RNA library in mice-bearing murine CT26 colon carcinoma liver metastases, identifying an aptamer that binds to p68, an RNA helicase upregulated in colorectal cancer. The same group has also further refined the developed strategy in order to develop a more useful reagent for human patients [71
]. As the target they used intrahepatic immunodeficient mice engrafted with two patient-derived cell lines (designated as 119X and 57X) from liver metastases. One of the selected molecules binds DHX9, an RNA helicase upregulated in colorectal cancer. As assessed by fluorescence molecular tomography imaging, this molecule is able to preferentially localize into the nucleus of the cancer cells in vivo following systemic administration, thus showing a potential use for targeted delivery. An additional in vivo selection strategy has been recently proposed to isolate aptamers able to cross the blood-brain barrier (BBB), the brain physiological barrier that protects the brain limiting at same time the therapeutic interventions in neurological disorders [72
]. The selection was developed by performing 14 rounds in which a 2 -fluoropyrimidine (2′-F-Py)-modified RNA library was administered to wild-type mice via tail vein injection and brains were harvested for aptamer recovery and amplification. Notably, among the isolated sequences, the authors found some aptamers able to enter brain endothelia and parenchymal cells after peripheral injection.
Despite their sophistication, these in vivo approaches have the great advantage that they can perform the selection in a natural physiological environment. In addition, aptamers are isolated based on their localization, permitting a “built in” negative selection, directly eliminating those sequences that disperse to organs/tissues not of interest.