Inverse In Vitro Selection Enables Comprehensive Analysis of Cross-Chiral L-Aptamer Interactions

Abstract

Aptamers composed of mirror-image L-(deoxy)ribose nucleic acids, referred to as L-aptamers, are a promising class of RNA-binding reagents. Yet, the selectivity of cross-chiral interactions between L-aptamers and their RNA targets remain poorly characterized, limiting the potential utility of this approach for applications in biological systems. Herein, we carried out the first comprehensive analysis of cross-chiral L-aptamer selectivity using a newly developed "inverse" in vitro selection approach that exploits the genetic nature of the D-RNA ligand. By employing a library of more than a million target-derived sequences, we determined the RNA sequence and structural preference of a model L-aptamer and revealed previously unidentified and potentially broad off-target RNA binding behaviors. These results provide valuable information for assessing the likelihood and consequences of potential off-target interactions and reveal strategies to mitigate these effects. Thus, inverse in vitro selection provides several opportunities to advance L-aptamer technology.

Keywords: In vitro selection; L-aptamers; RNA; selectivity; sequencing.

Figure 1.

Schematic representation of the inverse in vitro selection method and comparison to a standard selection approach. During inverse in vitro selection, the unhybridized loop domain of the RNA target, not the L-aptamer, undergoes selective enrichment to identify potential off-target sequences.

Figure 2.

Sequences and secondary structures of L-6-4t, D-TAR RNA, and library D-L6N. The boxed nucleotides in D-TAR indicate the L-6-4t binding site and were randomized to generate D-L6N. Sequences of all oligonucleotides are listed in Table S1.

Figure 3.

Saturation plots for binding of L-6-4t to D-TAR and D-TAR_UG. For simplicity, only the sequences of the 6-nt loop domain of the D-TAR hairpin is shown (boxed). The K_d is indicated in parenthesis. Data are mean ± S.D. (n = 3). Representative EMSA gels are depicted in Figures S3 and S6a.

Figure 4.

(a) Sequence and secondary structure of the D-L10N library. (b) Sequence logo depicting the relative frequency of nucleotides at each position within the random domain of D-L10N following 3-rounds of inverse selection (n = 70 sequences). (c) Saturation plots for binding of L-6-4t to D-10_UU, D-10_CC, and D-14. For simplicity, only the sequences of loop domain of the hairpin are shown (boxed). The K_d is indicated in parenthesis. Data are mean ± S.D. (n = 3). Representative EMSA gels are depicted in Figure S5.