High-Throughput Sequencing and SELEX-Based Protocol for Selecting Aptamers Against Potato Spindle Tuber Viroid

TL;DR

Researchers developed a new method using high-throughput sequencing and SELEX to select aptamers that bind to potato spindle tuber viroid, improving detection and analysis.

Contribution

A novel MB-HTS-SELEX protocol is introduced for selecting aptamers against viroids, which was previously unexplored.

Findings

A magnetic bead-based SELEX protocol was successfully used to select potential PSTVd-binding aptamers.

High-throughput sequencing identified enriched sequences after 10 selection rounds.

Pull-down assays confirmed that the most abundant oligo-ssDNA in L30 binds to PSTVd molecules.

Abstract

Aptamers are powerful tools for detecting and analyzing biomolecules that consist of proteins or nucleic acids. However, their application to aptamers against viroids—highly structured self-replicating RNAs—has not yet been explored. In this study, a magnetic bead- and high-throughput sequencing-based SELEX (MB-HTS-SELEX) protocol for selecting potential aptamers against potato spindle tuber viroid (PSTVd) is presented. Full-length biotinylated-PSTVd RNA was transcribed in vitro, immobilized on streptavidin-coated magnetic beads, and incubated with a library of ~3.32 × 1014 molecules of random single-stranded oligo-DNAs (oligo-ssDNAs) of 20, 30, or 40 nucleotides (L20, L30, or L40, respectively) flanked by primer binding sites for downstream PCR amplification. Simultaneous biotin labeling of the anti-aptamer strand of the resulting double-stranded DNA (dsDNA) amplicons facilitated strand separation using streptavidin-coated magnetic beads. After 10 selection rounds, high-throughput sequencing, followed by bioinformatics analysis of the generated sequences, allowed for the detection of several enriched sequences, representing putative PSTVd-binding aptamers. Subsequent pull-down assays showed that the most abundant oligo-ssDNA in L30 was docked on PSTVd molecules. This combination method may ameliorate the selection of high-affinity aptamers against PSTVd, reduce the number of selection cycles, time, and other costs of aptamer production, thereby promoting future massive and cost-effective viroid detection and characterization.

Figures

Full text

1. Introduction

Viroids are single-stranded, circular, and highly structured small RNA molecules (234–451 nucleotides) that can replicate based on RNA genome, but they lack protein-coding capacity and immunogenicity. They can cause diseases through the interaction of their RNA structure with plant host factors, enabling host exploitation for viroid replication, transport, circumvention of host defense, and alteration of host gene expression [1,2,3,4,5]. Among other important pathogens of solanaceous species, potato spindle tuber viroid (PSTVd, Pospiviroid fusituberis, family Pospiviroidae, genus Pospiviroid) may induce epinasty and stunting in tomato, as well as various malformations in potato. PSTVd, which is around 360 nucleotides in length, adopts a typical rod-like secondary structure, consisting of five structural domains, namely, terminal left (TL), pathogenicity (P), central (C), variable (V), and terminal right (TR) [6,7], which is a characteristic of all the members of the family Pospiviroidae [8,9]. This native rod-like molecule is a sequence of stems, where RNA is double stranded, and loops or bulges, where RNA is single stranded (Figure 1). This structural sequence is essential for PSTVd pathogenicity, including transport, replication, movement, and symptom expression [10]. In particular, features such as the thermolability or bent structure of the P-domain have been reported to play a key role in the severity of disease symptoms [11,12,13,14,15].

Oligonucleotide aptamers are nucleic acid (DNA or RNA) oligomers that bind specifically to a given ligand by folding around it, thereby forming a specific binding site, which resembles an antibody and its epitope on an antigen [16,17]. These aptamers rely on the same fundamental binding forces that confer affinity and specificity in antigen–antibody complexes, including hydrogen bonding, electrostatic complementarity and electrostatic interactions, π–π system stacking, and van der Waals forces that determine affinity and specificity in antigen–antibody complexes [18,19]. Other identities comprise a sequence with a length of 15–45 nucleotides, a binding affinity ranging from micromolar to picomolar scales, and discrimination among closely related ligands [20]. Aptamers are commonly produced by an evolutionary screening method known as the Systematic Evolution of Ligands by EXponential enrichment (SELEX). A typical SELEX procedure of around 10–15 cycles (2–4 days/cycle) may require up to 3 months to complete, which is half the time of the minimum 6 months required for monoclonal antibody development, yet it is still considered laborious and time consuming [18,20,21,22]. Aptamers can function as synthetic antibodies, which can detect molecules lacking protein coding capacity and subsequent immunogenicity, such as viroids.

To date, the production of aptamers for detecting viral proteins, viral RNAs, and ribozymes has been reported [23]. Notably, a DNA aptamer-based biochip has been designed to detect apple stem pitting virus (Foveavirus mali) coat proteins PSA-H and MT32 [24,25]. Selecting aptamers against viroids is important, as aptamers can substitute for antibodies, thereby allowing the use of aptamers for viroid diagnosis, especially in plant-health-certification schemes. In addition, by studying aptamer–viroid binding-site pairing, the subsequent aptamers that hinder viroids from replicating or interacting with plant elements can further reveal mechanisms elucidating viroid biology. Targeting PSTVd for aptamer selection is a challenging task, given that there are few chances for an oligonucleotide to interact with the rod-like structure of PSTVd, and that oligonucleotides cannot easily recognize negatively charged RNA structures containing phosphate groups [26,27]. However, an oligonucleotide could interact with unpaired bases in loops and bulges through non-Watson–Crick pairing or could form base triplets within a stem [26,28,29].

A thorough standardization of SELEX is indispensable for optimizing a system that selects strong binders against nucleic acid targets, adjusting a number of factors such as target and random oligonucleotide pool design, selection and partition environment, amplification protocol, number of selection rounds, and other factors affecting stringency [20,26]. For SELEX of single-stranded DNA (ssDNA) libraries, the conversion of amplified dsDNAs at the end of the selection round to ssDNA for the next round is an additional critical step, which determines the initial amount (i.e., diversity) of the selected binders subjected to every subsequent selection round [27,30,31,32]. Pre-SELEX and post-SELEX modifications, such as biotin or digoxigenin (DIG) labeling of the screened library for DNA strand separation, are methods of choice that allow for the efficient and easy recovery of aptamer strands [19,22,33,34,35].

High-throughput sequencing (HTS) involves a sequencing platform-based approach for producing and assembling millions of DNA and cDNA sequence reads, assisted by strong bioinformatics packages. HTS coupled with SELEX can facilitate anti-PSTVd aptamer production by significantly reducing the selection rounds and their costs and allowing for the quantitative assessment of the library during selection [36,37].

Post-MB-HTS-SELEX experiments such as pull-down assays can be used to determine the affinity between potential aptamers and a viroid target. Pull-down assays are similar to the opposite selection round, as the selected biotinylated oligonucleotide is immobilized on streptavidin-coated beads, and an unmodified target is incubated with the immobilized oligonucleotide in the presence of other molecules, for example, RNA molecules from a host plant, under selective binding conditions. Following incubation, non-binding molecules are removed through washing, and any viroid-oligonucleotide complexes are subsequently isolated via magnetic separation. Then, the binding fraction is released from the beads through elution and subjected to hybridization using a PSTVd-specific probe, in the presence of proper controls. Positive hybridization signals may confirm the binding between the tested oligonucleotide and the viroid.

In this study, a SELEX protocol coupled with HTS for selecting potential aptamers from single-stranded oligo-DNA (oligo-ssDNA) pools was established to facilitate rapid screening of possible anti-PSTVd sequences.

2. Results

2.1. Establishment of the SELEX System for PSTVd

2.1.1. PSTVd Target for SELEX

Using the full-length PSTVd DNA construct starting from the positions 180 to 179 (Figure 1A) as a template, in vitro transcription was performed with T7 RNA polymerase using the biotin RNA labeling mixture. The biotinylated transcripts with a length of about 360 nucleotides were visualized by 8 M urea–5% polyacrylamide gel electrophoresis (PAGE) (Figure 1B), from which they were eluted, dissolved in distilled water, and used for SELEX as the target.

2.1.2. Design of Random Oligo-ssDNA Libraries for Aptamer Selection by SELEX

To investigate whether DNA aptamers capable of binding PSTVd could be obtained, random oligo-ssDNA populations with the lengths of 20, 30, and 40 nucleotides were constructed. In performing the SELEX rounds to select aptamers, the selected oligo-ssDNAs bound to the target must be amplified using PCR, and then the aptamer strand must be separated from the amplified dsDNAs. Hence, two pairs of external and internal primer sequences were added to both ends of random oligo-ssDNAs, as well as the control oligonucleotides (weak binder Pw5 and strong binder PSTVd-P, see the Materials and Methods section), to avoid sequence similarity to the target PSTVd RNA, except for PSTVd-P. As shown in Table 1, two primers were finally designed: a pair of forward and reverse primers for the internal primers and a pair of M13 forward and T7 reverse primers for the external primers.

2.1.3. Optimal Mixture Ratio of Binder and Target with Magnetic Beads Used for Trapping and Elution

Among the magnetic beads tested, Dynabeads M270 and C1 were more efficient in trapping and releasing binders. The best selection conditions, specifically, the efficient trapping and releasing of the biotinylated-PSTVd coupled with aptamer candidates, were achieved on Dynabeads M270 and C1 using Invitrogen binding and washing (B&W) buffer at room temperature (20–30 °C). In addition, in a preliminary analysis using the strong binder PSTVd-P, the amount of the binder eluted from PSTVd targets bound to streptavidin-coated Dynabeads M270 reached saturation at the binder-to-PSTVd target ratio between 1:1 and 10:1 (Table 2), and was therefore set at 10:1. Preheating and denaturing the target and binders (or strong binder PSTVd-P) at 65 °C was indispensable for their proper folding prior to binding reactions.

2.1.4. Strand Separation

Conditions for strand separation were analyzed using the L30 library after four selection rounds. Briefly, an aliquot of the selected oligo-ssDNAs was subjected to PCR using a biotinylated T7-Internal reverse and DIG-labeled M13-Internal forward primer set, and amplified dsDNAs labeled with DIG for the aptamer strand and biotin for the anti-aptamer strand. The amplicon was incubated with streptavidin-coated magnetic beads in B&W buffer and then separated via incubation in 1× SSC at 95 °C for 5 min to release the non-biotinylated strand into the supernatant. Following eluate recovery, the bead fraction was further incubated in 0.15 M NaOH at room temperature for 15 min to release the remaining non-biotinylated strand. Both eluates obtained via heat and NaOH treatments were blotted onto a nylon membrane and subjected to DIG-ELISA. As shown in Figure 2, both showed positive signals. Therefore, the combination of heat treatment with 1× SSC at 95 °C for 5 min and chemical denaturation in 0.15 M NaOH efficiently recovered the aptamer strand via strand separation.

2.2. Selection of PSTVd-Binding Aptamers via SELEX

2.2.1. Selection Against Full-Length PSTVd

Based on the set conditions (repetitions, temperature, and reaction volume) for binding, washing, and elution presented in Table 3, the SELEX reaction was initiated using the L20, L30, and L40 random oligo-ssDNA libraries as binders and the full-length PSTVd transcript as the target. Interim analysis of the eluate after the fourth selection round showed that the expected size of amplicons was delivered by PCR using a M13-Internal forward and T7-Internal reverse primer set, confirming that the design of the L20, L30, and L40 random oligo-ssDNA libraries that serve as binders, the magnetic beads used, and the conditions for amplifying the selected binders by PCR worked as anticipated. The selection was performed for up to 10 rounds, with gradually increasing binding and washing stringencies, as summarized in Table 3. Negative selection was performed prior to rounds 1, 5, and 10.

2.2.2. HTS-SELEX

HTS by NGS of the oligonucleotide mixture amplified from the L20, L30, and L40 libraries selected through 10 rounds of SELEX, in addition to the similarly treated strong and weak binders, indicated that the proportion of High Quality Bases with ≥Q30 in the total number of bases after quality filtering was 95.62%, with a mean quality score of 36.76. After trimming the adapter sequences, they were sorted by size: L20 (19–21 nucleotides), strong binder (26–28), weak binder and L30 (29–31), and L40 (39–41). Given that the sequences of the strong and weak binders are known, similar sequences were also included in the analysis by accounting for the sequence information. Furthermore, from them, read counts of ≥30 were selected for result evaluation. Of the 10,982,726 reads obtained, the top 9957 sequences with the read count greater than 30 accounted for 2,518,511 reads (Supplementary Table S1). The sequences that perfectly matched the strong and weak binders accounted for 633,145 and 63,826 reads, which represented 82.70% and 84.70% of the total read in these groups, respectively. The various other derivatives found in these populations are thought to have emerged during repeated PCR amplification. This is because each SELEX experiment was initiated from a single strong- or weak-binder sequence and involved repeated PCR amplification over 10 rounds. Therefore, the minor variants observed, including indels and substitutions involving one to a few nucleotides, are likely attributable to PCR-induced errors (Table S1). On the other hand, as expected, the populations selected from the L20, L30, and L40 libraries contained diverse sequences. The top 10 predominant sequences in the L20, L30, and L40 libraries are shown in Table 4. Among them, the top three sequences in L30 (i.e., L30.10R.1, L30.10R.2, and L30.10R.3) were enriched significantly, accounting for 13.64%, 11.77%, and 1.63% of the population, respectively. In contrast, no significantly dominant sequences were found in the L20 and L40 populations, with the highest being 0.84% in L40 and only 0.15% in L20 (Table 4).

2.3. Binding Validation of Selected Aptamers to PSTVd

Pull-Down Assays

The affinity of the selected oligo-ssDNAs to PSTVd was examined by using a pull-down assay. In this assay, the biotinylated core sequence of the selected oligo-ssDNAs lacking the primer binding regions, L30.10R.1, L30.10R.2, and L30.10R.3, the top 50th oligo-ssDNA from the L30 library (i.e., L30.10R.50, 5′-TGAGGTGGTAAGACTAGCTGTTCGTCGTTTCGGG-3′), T7 reverse primer-like sequence (5′-TCCTTCTCTTCCCCTATAGTGAGTCGTATTAAGATCGGAA-3′), and the sequence reverse to the weak binder Pw5 (L30.rev: 5′-ATCAACAGGCCACCCACCGGGCGAGGAAGC-3′), which was used as a negative control, were custom synthesized. Then, they were incubated with a full-length PSTVd transcript, bound to streptavidin-coated magnetic beads, washed, and eluted PSTVd being trapped by the selected oligo-ssDNAs under the same conditions as the 10th round of SELEX.

A portion of the eluted fraction was diluted by fivefold, then dot blotted on a nylon membrane, and hybridized with the DIG-labeled PSTVd cRNA probe. The eluate from L30.10R.1 provided strong signals down to 1/625 dilutions, which was similar to the positive control (PSTVd transcripts, 1 ng), indicating the strong binding affinity of L30.10R.1 to PSTVd (Figure 3).

Another portion of the same eluted fraction was further examined by gel-blot Northern hybridization assay using the DIG-labeled PSTVd cRNA probe. The result indicated that a clear band corresponding to the full-length PSTVd RNA was visible in the lane of L30.10R.1 (Figure 4, arrow), thereby confirming the presence of PSTVd complexed with the L30.10R.1.

3. Discussion

Viroids are the smallest infectious RNA pathogens that replicate in host cells and incite specific diseases in some important crop plants. At present, the method that can effectively detect viroids is based on nucleotide sequence similarity, such as molecular hybridization and PCR. However, in studying viroids, which form highly ordered molecular structures to express complex functions within host cells, the development of molecules (e.g., nucleic acid aptamers) that recognize and interfere with highly ordered viroid structures has long been anticipated. To that end, the optimization of selection conditions is critical prior to actual selection using an actual library of potential aptamers [18,20,27]. In this study, SELEX was combined with magnetic beads and HTS technologies to examine various conditions for developing nucleic acid aptamers that interact with viroids. The factors examined ranged from the size of the random oligo-ssDNA library to the ratios of the binders to target, washing and elution conditions, types of magnetic beads, and detection methods for the selected oligo-ssDNAs (Figure 5).

Strand separation after recovering and amplifying the target-binding oligo-ssDNAs through SELEX is also a crucial step in preparing binders for use in the next round. Simultaneous biotin labeling of the antisense strands enabled the strand separation of purified dsDNA amplicons by binding the biotinylated anti-aptamer strand to streptavidin-coated magnetic beads [27,37]. In addition, treatments such as heating (at 95 °C for 5 min) and alkali (0.15 M NaOH) can effectively enhance the separation process and yield a better oligo-ssDNA concentration. However, this strand separation method is tedious; thus, further improvement is necessary.

In this study, HTS analysis of the L20, L30, and L40 libraries selected through 10 rounds of SELEX identified three enriched oligo-ssDNAs in L30. In contrast, the selections from L20 and L40 libraries showed lower enrichment, with lower abundance in the population than those from the L30 library, indicating that aptamers specific to PSTVd could not be effectively selected at lengths of 20 or 40 nucleotides. These results suggest that an aptamer length of approximately 30 nucleotides is suitable for viroid binding; however, this possibility has not yet been fully explored. Aptamer selection may also depend on compatibility with the primers or linkers added to both ends of the random sequence. Based on these results, further investigations focusing on the L30 library would be useful in subsequent studies.

HTS-SELEX, in general, can reveal significant similarities among random libraries subjected to one or several selection rounds. In several cases, HTS has been used to screen libraries for binding sites on specific targets after a limited number of selection rounds [37,38]. However, it has been reported that the most abundant oligo-ssDNAs in a library subjected to multiple selection rounds are not always the best binders on the target; therefore, additional evaluation, such as quantitative assessment of library composition, is necessary [36].

Therefore, the binding ability of the 30 nucleotides of oligo-ssDNAs, which became dominant after 10 rounds of SELEX, that is, L30.10R.1, L30.10R.2, and L30.10R.3, was indirectly assessed by using pull-down assays followed by dot-blot and gel-blot Northern hybridization. L30.10R.1 showed a positive pull-down activity, indicating its high binding affinity to PSTVd RNA (Table 5). To our knowledge, this study is the first to report on the successful selection of a DNA aptamer for a viroid.

As described, the top three oligo-ssDNAs shown in Table 5 bear no significant sequence similarity to PSTVd; however, one of them showed binding capacity to in vitro-transcribed PSTVd. Therefore, the binding capacity of the selected aptamers needs to be further addressed by calculating their dissociation constants (Kd) in future studies. Additionally, in order to use the selected aptamers in practical applications such as pull-down assays or enzyme-linked aptamer adsorption assays, it is necessary to examine their specific binding to PSTVd using purified native PSTVd and crude extracts from infected plants. Furthermore, analyzing the affinity of the aptamers to other viroid species belonging to the genus Pospiviroid, not only to PSTVd used in this experiment, is indispensable for the practical application of aptamer-based viroid disease diagnosis or protection in the future.

The protocols of MB-HTS-SELEX, using biotin-labeled oligo-ssDNAs (modified from [39,40]), are currently being further optimized to produce more sustainable results. HTS has been shown to reduce the time and the costs required for selection and to improve the efficiency of SELEX. Combining HTS with rigorous selection protocols can lead to more efficient selection of aptamers with diverse properties. In this study, the primary objective was to verify the feasibility of obtaining ssDNA aptamers capable of binding viroids. Therefore, SELEX conditions were examined using random oligo-ssDNA populations of 20, 30, and 40 nucleotides in length. The results demonstrated that it was possible to select oligo-ssDNAs that bind to PSTVd. However, repeated validation for each size was not completed. Therefore, it is important to conduct repeated experiments under the same conditions, such as using the L30 library, to further explore SELEX conditions for more reliable aptamer selection.

Finally, the results obtained in this study can lead to the cost-effective detection of viroids with aptamer-based assays in the near future. Furthermore, utilizing aptamers for viroid research to gain a comprehensive understanding of the structure, pathogenicity, and biology of viroids can contribute to the development of innovative means of viroid control.

4. Materials and Methods

4.1. PSTVd Target

A synthetic full-length PSTVd-intermediate isolate (GenBank accession number: M16826) [8] starting from the nucleotide at the position 180 U and ending at 179 U (i.e., 359 nucleotides in length) placed downstream of the promoter sequence of T7 RNA polymerase (Figure S1) was custom synthesized by Hokkaido System Science Co., Ltd. (Sapporo, Japan) and cloned into the pUC19 vector at the NotI site, which was used to transform Escherichia coli DH5α competent cells (Takara Bio Inc. Kusatsu, Japan). The selected E. coli colonies were further propagated in Luria–Bertani medium and used to extract the recombinant plasmid for in vitro transcription. The full-length biotinylated-PSTVd was transcribed by T7 RNA polymerase (Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA, USA) using the biotin RNA labeling mixture (Roche Diagnostics, Basel, Switzerland) as the substrate in accordance with the manufacturer’s instructions. After transcription, the PSTVd transcript was denatured at 68 °C for 15 min in the presence of 50% urea and electrophoresed by 8 M urea–5% PAGE (acrylamide:bisacrylamide = 39:1) at 450 V, 45 mA for 45 min in 1× TBE buffer [89 mM Tris-borate; 2 mM EDTA]. The gel slice containing the PSTVd transcript was excised from the gel, put in gel elution buffer [0.5 M ammonium acetate, 1 mM EDTA, and 0.1% SDS], and incubated overnight at 37 °C with shaking to allow for elution. After ethanol precipitation, the PSTVd transcript was dissolved in sterile distilled water, and used as the target in the SELEX experiment. The PSTVd transcript for pull-down assays was prepared similarly but was used without biotinylation.

4.2. Random Oligo-ssDNA Libraries

Three random oligo-ssDNA libraries comprising a random 20-, 30- or 40-nucleotide sequence (L20, L30, and L40, respectively) flanked by 40- and 39-nucleotide primer binding sequences on the left and right, respectively, were custom synthesized by Hokkaido System Science Co. Ltd. (Table 1).

4.3. Strong and Weak Binders Used as Controls

A DNA sequence complementary to the nucleotide positions from 47 to 73, which is 27 nucleotides in length, in the upper strand of the P-domain of PSTVd was custom synthesized and used as a strong binder (Table 1; PSTVd-P). The sequence was flanked by the same primer binding sequences as the L20, L30, and L40 random oligo-ssDNA libraries.

A weak binder was selected from the L30 library. That is, the L30 library was mixed with the full-length PSTVd target and was subjected to one selection round under the conditions presented in Table 3. An aliquot of the washed-away part of the library (i.e., the fraction that does not bind to the target) was recovered, PCR amplified, and cloned into the pGEM-T vector (Promega) for the transformation of E. coli DH5a competent cells (Takara Bio Inc.). The successfully cloned binders were sent to Solgent Co., Ltd., Daejeon, Korea for Sanger sequencing. Among the 20 sequences obtained, the most dissimilar to the strong binder PSTVd-P with regard to sequence was selected as the weak binder (Pw5; Table 1).

4.4. Magnetic Beads

After testing several kinds of streptavidin-coated magnetic beads, that is, Dynabeads™ MyOne™ Streptavidin C1 (Dynabeads C1), Dynabeads™ MyOne™ Streptavidin T1, Dynabeads™ M270 Streptavidin (Dynabeads M270) (Invitrogen), MagCapture™ Tamavidin^®^ 2REV (FUJIFILM Wako Pure Chemical Corp., Osaka, Japan), and Streptavidin MagneSphere^®^ Paramagnetic Particles (SA-PMPs) (Promega, Madison, WI, USA), Dynabeads M270 and Dynabeads C1 with a binding capacity of 200 and 500 pmol of biotinylated oligonucleotides/mg, respectively, were used for binding to the biotinylated-PSTVd targets in the selection rounds of SELEX. On the other hand, both Dynabeads M270 and SA-PMPs, mainly the latter, were used for strand separation. Both results were comparable.

4.5. Buffers for Binding, Washing, and Elution

Binder–target binding buffer [10 mM Tris-HCl, pH 7.5; 1 mM EDTA; and 150 mM NaCl] was used for binding the binder (random oligo-ssDNAs) and target (biotinylated-PSTVd). B&W buffer [5 mM Tris-HCl, pH 7.5; 0.5 mM EDTA; and 1 M NaCl] was used for binding and washing of the biotinylated-PSTVd–binder complex and streptavidin-coated magnetic beads according to the instructions for Dynabeads™ M270 Streptavidin supplied by Invitrogen. Binders bound to the target were recovered using elution buffer E [10 mM Tris-HCl, pH 7.5; 1 mM EDTA; and 25 mM NaCl]. For strand separation, 1× SSC buffer was used, prepared by dilution of a 20× SSC stock solution (3 M NaCl and 0.3 M sodium citrate, pH 7.0). TE buffer was prepared as a 10× stock solution (100 mM Tris-HCl, pH 7.5; 10 mM EDTA) and diluted as required.

4.6. Selection Against Full-Length PSTVd

For SELEX against PSTVd, the L20, L30 or L40 random oligo-ssDNA library (~550 pmol) was first incubated with the full-length PSTVd transcript (~55 pmol) in binder–target binding buffer. The resulting complexes were then captured on streptavidin-coated magnetic beads in B&W buffer. After incubation, the non-specifically bound oligo-ssDNAs were washed away with B&W buffer, and the bound fraction was eluted with buffer E. The incubation temperature, time, and repetition for each step are presented in Table 3. Namely, the selection was strictly pressure-driven by increasing stringency from round to round. Neutral SELEX was not considered, i.e., neither repeated PCR amplification with nor without binding to the target under low-pressure conditions. Strong binder PSTVd-P and weak binder Pw5 were similarly treated up to 10 selection rounds.

Negative selection was performed to remove oligo-ssDNAs that bind nonspecifically to non-target substances, such as beads or tubes. Random oligo-ssDNA libraries were first heated at 95 °C for 1 min and immediately cooled to 4 °C for 1 min allow folding. Then, 0.06 mL B&W buffer containing 0.6 mg streptavidin-coated Dynabeads^TM^ M270 or C1 was added to the respective libraries, followed by incubation at room temperature for 15 min with gentle pipetting twice at 7.5 min intervals. After magnetic separation for 2 min, the supernatant was recovered, an equal volume of B&W buffer was added to the beads, and the mixture was incubated again for 15 min with mixing at room temperature. After a second magnetic separation for 2 min, the supernatant was recovered and used for SELEX in the next round. Negative selection was performed prior to rounds 1, 5, and 10.

4.7. PCR Amplification of the Selected Oligo-ssDNAs and Gel Purification of the Amplified DNAs

The selected oligo-ssDNAs were amplified via PCR in a 50 μL reaction solution [1× Quick Taq^®^ HSDyeMix (Toyobo Co. Ltd., Osaka, Japan); 0.4 M trimethylsulfoxonium (TMSO); and 0.2 µM of each of M13-Internal forward and biotinylated T7-Internal reverse primers] containing 2 μL of the selected oligo-ssDNAs, with the parameters of an initial denaturation step (95 °C, 5 min), 35 cycles of amplification consisting of denaturation (95 °C, 30 s)–annealing (68 °C, 30 s)–extension (68 °C, 30 s), and a final extension step (68 °C, 5 min). An aliquot of the amplicon was electrophoresed on 10% PAGE, and the bands were visualized by silver staining to confirm the amplification. Then the biotinylated amplicons were purified by 3% agarose gel electrophoresis (AGE) in TAE buffer [40 mM Tris-acetate; 1 mM EDTA], and the band excised from the gel was immersed in elution buffer E and shaken overnight to recover the amplification products. After ethanol precipitation, the products were dissolved in ~20 μL of sterile distilled water.

4.8. Strand Separation

To continue SELEX for multiple rounds, it is essential to amplify the selected oligo-ssDNAs via PCR for use subsequent rounds (Figure 5b). Because PCR produces dsDNA, the conversion of amplified dsDNA to ssDNA, referred to as strand separation, is a critical step. For this purpose, a portion of the gel-purified PCR product containing the biotinylated anti-aptamer strand (Figure 5c) was mixed with streptavidin-coated magnetic beads (Magnesphere SA-PMPs) in B&W buffer and incubated at room temperature for at least 30 min. The mixture was then placed on a magnetic stand at room temperature for 3 min to remove unbound biotinylated dsDNA amplicons. The beads were washed twice with B&W buffer and incubated in 1× SSC at 95 °C for 5 min to release the non-biotinylated aptamer strand (Figure 5d). After magnetic separation, the eluate (i.e., non-biotinylated aptamer strand) in 1× SSC was recovered, and the beads were further incubated in 0.15 M NaOH at room temperature for 15 min to release the remaining aptamer strand (Figure 5d). The NaOH eluate recovered was titrated with 10× TE (pH 7.5)–1.25 M acetic acid, ethanol precipitated, and reconstituted in sterile distilled water.

At the end of the round, the recovered oligo-ssDNAs (i.e., aptamer strand recovered by strand separation, Figure 5e) were preheated at 95 °C for 8 min, cooled at 4 °C for 3 min allow folding, and adjusted to room temperature by adding an equal volume of 2× B&W buffer. To start the next selection round, the folded oligo-ssDNAs were mixed with the biotinylated full-length PSTVd bound on Dynabeads^TM^ M-270 in 1× B&W buffer (Figure 5a). This procedure was repeated for 10 rounds under gradually increasing selection stringency, including binding temperature, washing temperature, and the incubation time and frequency (Table 3).

4.9. HTS-SELEX

The oligo-ssDNAs recovered from libraries L20, L30, and L40 after 10 selection rounds were amplified, respectively, in a 50 μL of PCR mixture using the TaKaRa LA Taq^®^ Hot Start Version (Takara Bio Inc.) consisting of 1× HS LA Taq Buffer II, 1 unit LA Taq^®^ HS, 0.25 mM each dNTP, 2.5 mM MgCl_2_, 0.2 µM each of the primers M13-Internal forward primer and T7-Internal reverse primer, and 1 μL selection eluate, with an initial denaturation step (95 °C, 10 min), 35 cycles of amplification consisting of denaturation (95 °C, 30 s)–annealing (68 °C, 30 s)–extension (72 °C, 30 s), and a final extension step (72 °C, 5 min). These amplicons were analyzed on 10% PAGE and visualized by silver staining [41,42]. Then, they were purified by 3% AGE in TAE buffer. The concentration of the purified dsDNA amplicons was estimated using Nanodrop Microvolume Spectrophotometer (Thermo Fisher Scientific Inc.), and equal amounts of the purified dsDNA amplicons were mixed before sending to Hokkaido System Science Co. Ltd. for HTS analysis using the Hiseq Paired-End method (100 bp reads) in accordance with the Hiseq Illumina RUN protocol.

The resulting reads were analyzed using tools on the Galaxy platform for joining the paired-reads, Clustal alignments, and collapsing sequences (i.e., counting sequence frequency in the library) (https://usegalaxy.org/root, accessed on 11 February 2026) [43]; and, Geneious v. 10.1.3 for Windows [44] for trimming the adapter and primer binding regions.

4.10. Pull-Down Assay Followed by Dot-Blot or Gel-Blot Northern Hybridization

The sequences that appeared at high frequency in the HTS analysis were subsequently subjected to pull-down assays to investigate their binding capacity to PSTVd. The short 5′-biotinylated version of oligo-ssDNAs, corresponding to a random region (30 nucleotides in length), was custom synthesized (Hokkaido System Science Co. Ltd.). Approximately 7 pmol of biotinylated oligo-ssDNAs L30.10R.1, L30.10R.2, and L30.10R.3, which were among the most enriched sequences in the 10th selection round of the L30 libraries, were immobilized on streptavidin-coated Dynabeads M270 (Invitrogen) and incubated with 112 ng (approximately 1 pmol) of full-length PSTVd transcript, under the same conditions as in the 10th selection round.

The bound fraction was collected, and then an aliquot was denatured in a mixture of 50% formamide, 6.5% formaldehyde and 0.5× SSC, at 68 °C for 15 min, transferred immediately on ice, and complemented with an equal volume of 20× SSC. Next, 1, 1/5, 1/25, 1/125, and 1/625 dilutions were made in 10× SSC buffer, and each dilution was dot-blotted directly onto a positively charged nylon membrane (Hybond™ N+, Amersham, Cytiva, Marlborough, MA, USA). The membrane was also spotted with 1 ng of PSTVd in the same dilutions.

The membrane was cross-linked at 1200 × 100 µJ/cm^2^, prehybridized at 55 °C for 1 h, and incubated overnight at 55 °C in a hybridization solution (50% formamide; 10% dextran sulfate; 0.18 M NaCl; 20 mM sodium phosphate, pH 7.0; 0.1% SDS; 500 μg/mL yeast t-RNA; and 25 µg/mL sonicated salmon sperm DNA) containing ~600 ng of the DIG-labeled PSTVd cRNA probe per ~50–100 cm^2^ of membrane [45]. DIG-labeled PSTVd cRNA probe was transcribed from recombinant pBlueScript II (SK-) (Agilent Technologies, Inc., Santa Clara, CA, USA) containing dimeric full-length PSTVd-dahlia isolate (GenBank accession number: AB623143) by T7 RNA polymerase (Invitorogen, Thermo Fisher Scientific) using DIG RNA labeling mix (Roche Diagnostics) as a substrate. After washes with 2× SSC at room temperature, RNase A digestion in 2 × SSC, and other washes with 0.1× SSC–0.1% SDS at 65 °C, the membranes were submerged in a blocking solution containing blocking reagents (Roche Diagnostics) in maleic acid buffer [0.1 M maleic acid, pH 7.5; 0.15 M NaCl] and then incubated with 1/7500 dilution of anti-DIG Fab fragment (Roche diagnostic). After two washes with washing buffer [0.1 M maleic acid, pH 7.5; 0.15 M NaCl; 0.3% Tween 20], the membranes were soaked in a developing solution consisting of 200 μL of CSPD-star (Roche diagnostic) in 500 μL of buffer 3 [0.1 M Tris-HCl, pH 9.5; 0.1 M NaCl; and 0.05 M MgCl_2_] [46] and then exposed for 1 h in a ChemiDoc XRS system (Bio-Rad Laboratories, Inc., Hercules, CA, USA).

To confirm the dot-blot findings, an aliquot of eluates in the pull-down assay was dissolved in 50% urea solution, denatured at 68 °C for 15 min, charged on 8 M urea–7.5% PAGE in 1× TBE buffer, and then electrophoresed at 450 V, 45 mA for 30 min. The gel-fractionated RNAs were transferred to a positively charged nylon membrane (Hybond™ N+, Amersham) by contact blotting overnight at 25 °C, cross-linked, and hybridized with the DIG-labeled PSTVd cRNA probe. After washing and DIG-ELISA as above, the membrane was exposed for 1 h in a ChemiDoc XRS system.