Cell-SELEX and Counter-SELEX

The overall procedure of the modified Cell-SELEX is schematically presented in Scheme 1. First, 20 nmol of the initial ssDNA library was dissolved in 1000 μL of the binding buffer (5 mM MgCl₂, 4.5 g of glucose, 1 g BSA, and 100 mg yeast tRNA in 1 L Dulbecco’s PBS), followed by heating at 95 °C for 5 min and then snap-cooling on ice. The refolded pool was then incubated directly to over 5 million HeLa^ASPH cells for 1 h. After washing, the bound sequences were recovered and amplified via 10 cycles of PCR (hot start: 2.5 min, 95 °C; denaturation: 0.5 min, 94 °C; annealing: 0.5 min, 46 °C; extension: 0.5 min, 72 °C; final extension: 5 min, 72 °C). An additional PCR was carried out at the cycles of 4, 6, 8, 10, and 12 to determine the optimum number of cycles for a preparative PCR. Single stranding of the products was done using lambda exonuclease III (Thermo Fisher Scientific) according to the manufacturer’s protocols. The counter selection was initiated from the second round when the recovered ssDNA from the positive selection was renatured and then incubated with the control untransfected HeLa. To enhance the affinity of the selected aptamers, the number of washing as well as the duration and volume were gradually increased, while the incubation time was reduced to 30 min. Meanwhile, the cell number was gradually reduced to 1 million in cycle 5. Moreover, up to 20%, FBS was added gradually to the binding buffer (Table S1). After the last round of selection (round 9), the PCR product was sequenced by GenXPro GmbH (Frankfurt, Germany) using Illumina NextSeq. 500 (1 million reads, 1 × 75 bps).

To overcome the imperfectness of the counter selection and to increase the chances of recognizing and eliminating the unduly retained oligomers, we effectively established a novel negative selection procedure named Counter-SELEX. In this method, the oligomers bound to the control cells at the second round of selection were literally recovered, amplified, and then subjected to five iterative rounds of selection, using control (untransfected) HeLa cells as the main target of SELEX (Table S2). Finally, using deep sequencing, the most prevalent sequences bound to the common surface molecules on the control cells were determined.