An RNA aptamer that shifts the reduction potential of metabolic cofactors

The study investigates whether RNA can employ differential affinity for oxidized versus reduced redox cofactors to modulate thermodynamic properties, specifically midpoint reduction potentials (Em), akin to protein oxidoreductases. Given RNA’s known catalytic roles and its ability to form specific ligand-binding pockets (e.g., riboswitch aptamers), the authors hypothesize that RNA could shift cofactor Em through selective recognition of oxidation states, thereby broadening the scope of RNA-mediated redox chemistry in a putative RNA world and enabling synthetic biology applications. The context includes the centrality of redox reactions in metabolism, the prevalence of adenylated cofactors (NAD+, FAD), and the wide Em tuning by flavoenzymes via local pocket environments. Demonstrating similar Em tuning by RNA would support plausible RNA-world metabolisms and provide design principles for RNA-based oxidoreductases.

Prior selections identified aptamers for nicotinamides and flavins, but oxidation-state preference was generally not enforced. A reported NAD+ aptamer showed some oxidation-state preference, without Em measurements. DNA aptamers for PQQ or for NAD+ adenosine did not alter Em. An NAD+-dependent alcohol dehydrogenase ribozyme modulated oxidation/reduction rates (kinetics) but no Em changes (thermodynamics). A prior RNA (Ftest1) recognized the FAD isoalloxazine ring without discriminating oxidation state. These works indicate that differential cofactor recognition and Em modulation are not generic properties of nucleic acids and require specific binding pocket features, motivating the present study.

• In vitro selection: RNA pool (N12 diversity) selected on immobilized FAD resin with ADP counter-selection to deplete adenosine binders. Increasing stringency over 12 rounds with mutagenic PCR at rounds 7–8. Hits truncated to B2, then stabilized to 38-nt X2B2.
• UV-Vis spectroscopy: Monitored ligand-induced spectral shifts of flavins (FAD, FMN, riboflavin) upon RNA binding in TKNM buffer. Denaturation or nonbinding RNA controls used.
• Metal ion dependence: Fluorescence quenching assays with FMN and various divalent cations (Mg2+, Mn2+, Zn2+, Ca2+; Co(NH3)6 3+ as negative control) to assess ion requirements.
• Mutational analysis: Systematic mutations in predicted stems (P1, P2) and loops (L1, L2); identification of X2B2-C14U with enhanced binding and larger spectral shifts. Additional mutagenesis of base-triple elements and position 14 substitutions/deletion; deoxyribose substitutions at G13 and/or U18 to probe 2′-OH roles.
• Isothermal titration calorimetry (ITC): Determined thermodynamic parameters and Kd values for RNA binding to FAD, FMN, riboflavin; exothermic interactions with micromolar–nanomolar affinities.
• Redox potential measurements: Xanthine oxidase-coupled assay with reference dyes (AQS, Em = −225 mV; phenosafranin, Em = −252 mV) under anaerobic conditions. Time-resolved UV-Vis to quantify oxidized/reduced ratios; Nernst-based linear fits to derive Em of free and RNA-bound flavins.
• NMR spectroscopy and structure determination: 1D/2D NMR (NOESY; imino proton spectra) with isotope-labeled RNAs to assign interactions; structure of X2B2-C14U–FMN derived using NMR distance restraints, CYANA, and Amber MD simulated annealing. Analysis revealed stacking and donor atom–π interactions, base-triple platform, and solvent accessibility of reactive FMN atoms.

• Discovery and minimization: Selection yielded FAD-binding aptamers; a 38-nt aptamer (X2B2) preferentially bound oxidized flavins over reduced forms (e.g., FAD vs FADH2).
• Spectral signatures: RNA binding red-shifted flavin absorption (FAD λmax: free 450 nm to 456 nm with X2B2; to 458 nm with X2B2-C14U) and introduced shoulders (+26 nm), consistent with altered hydrogen bonding environment.
• Metal ion requirements: Binding supported by Mg2+, Mn2+, Zn2+ (and Ca2+ for X2B2-C14U); not supported by Co(NH3)6 3+, indicating at least one inner-sphere contact with partially dehydrated metal ion.
• Affinities: ITC showed exothermic binding with micromolar–nanomolar Kd; strongest measured affinity Kd = 243 ± 28 nM for X2B2-C14U–FMN.
• Em shifts (mV; mean ± s.d., n=3): • Free flavins: FAD −209 ± 1; FMN −211 ± 2; riboflavin −210 ± 1. • Controls: Non-binding RNA + FAD −209 ± 1; Ftest1 + FAD −212 ± 3 (no substantial change). • X2B2 complexes: FAD −223 ± 1 (ΔEm = −11); FMN −234 ± 3 (ΔEm = −23); riboflavin −223 ± 1 (ΔEm = −14). • X2B2-C14U complexes: FAD −234 ± 1 (ΔEm = −25); FMN −251 ± 2 (ΔEm = −40; largest shift); riboflavin −234 ± 1 (ΔEm = −25).
• Structural mechanism (NMR/MD): FMN isoalloxazine xylene ring is π-π stacked between U14 (si-face) and G15 (re-face); multi-layer base-triple platform positions the ligand; donor atom–π contacts (e.g., U14 O4) contribute. The uracil edge of FMN does not engage in Watson–Crick/Hoogsteen pairing, consistent with spectral shoulders. Reactive FMN atoms C4a and N5 remain solvent-accessible.
• Mutational insights: Base-triple U30:A10:A9 critical for binding (disruption strongly reduces binding). Position 14 is pivotal: pyrimidine required; U14 enhances binding and yields more negative ΔEm than C14, consistent with donor atom–π effects (O4 vs N4). Deoxyribose substitutions at G13 and/or U18 reduce magnitude of ΔEm (e.g., dG13: −243 mV; dU18: −225 mV; double: −223 mV), suggesting altered stacking from sugar pucker changes.
• Mechanistic implication: Differential affinity favors oxidized flavin and disfavors reduced form, shifting Em negatively by up to ~40 mV, comparable to shifts in some flavoenzymes.

Findings demonstrate that RNA binding pockets can modulate the thermodynamic properties of bound redox cofactors by differential recognition of oxidation states, altering Em and thus the driving force for electron transfer. π-π stacking and donor atom–π interactions within the X2B2/X2B2-C14U pockets, and absence of hydrogen bonding to the FMN uracil edge, collectively favor oxidized flavin and disfavor the reduced form, producing negative Em shifts. The magnitude of the shifts, particularly for X2B2-C14U–FMN (−40 mV), overlaps with values observed in protein flavoenzymes (e.g., glutathione reductase), underscoring that RNA can achieve protein-like tuning through local electrostatics and stacking. These results support the plausibility of RNA-world metabolic redox chemistry and furnish design principles for engineering RNA-based oxidoreductase catalysts with tailored Em to broaden substrate redox windows.

The study identifies and characterizes a compact RNA aptamer (X2B2 and mutant X2B2-C14U) that preferentially binds oxidized flavins and shifts their midpoint reduction potentials by up to −40 mV. NMR-derived structural insights reveal π-π stacking and donor atom–π interactions as key determinants of differential affinity and Em tuning. This establishes RNA as capable of modulating cofactor thermodynamics similarly to proteins, suggesting mechanisms for ancient RNA-based metabolism and providing a basis to engineer RNA oxidoreductase catalysts in synthetic biology. Future work may leverage binding-pocket modifications to further tune Em, integrate these aptamers into catalytic constructs, and explore substrate-specific redox transformations.

• Functional catalysis was not demonstrated; the work establishes Em modulation and binding, not RNA-catalyzed redox reactions.
• Structural determination focused on the X2B2-C14U–FMN complex; generalization to other cofactors and variants relies on spectroscopic/ITC evidence rather than full structures.
• Divalent metal dependence and requirement for inner-sphere coordination may constrain environmental conditions for activity.
• Some measurements (e.g., deoxyribose mutants’ Em) had limited replicates (n=2) due to sample availability, potentially reducing statistical confidence for those specific comparisons.