Translesion synthesis by AMV, HIV, and MMLVreverse transcriptases using RNA templates containing inosine, guanosine, and their 8-oxo-7,8-dihydropurine derivatives

Abstract

Inosine is ubiquitous and essential in many biological processes, including RNA-editing. In addition, oxidative stress on RNA has been a topic of increasing interest due, in part, to its potential role in the development/progression of disease. In this work we probed the ability of three reverse transcriptases (RTs) to catalyze the synthesis of cDNA in the presence of RNA templates containing inosine (I), 8-oxo-7,8-dihydroinosine (8oxo-I), guanosine (G), or 8-oxo-7,8-dihydroguanosine (8-oxoG), and explored the impact that these purine derivatives have as a function of position. To this end, we used 29-mers of RNA (as template) containing the modifications at position-18 and reverse transcribed DNA using 17-mers, 18-mers, or 19-mers (as primers). Generally reactivity of the viral RTs, AMV / HIV / MMLV, towards cDNA synthesis was similar for templates containing G or I as well as for those with 8-oxoG or 8-oxoI. Notable differences are: 1) the use of 18-mers of DNA (to explore cDNA synthesis past the lesion/modification) led to inhibition of DNA elongation in cases where a G:dA wobble pair was present, while the presence of I, 8-oxoI, or 8-oxoG led to full synthesis of the corresponding cDNA, with the latter two displaying a more efficient process; 2) HIV RT is more sensitive to modified base pairs in the vicinity of cDNA synthesis; and 3) the presence of a modification two positions away from transcription initiation has an adverse impact on the overall process. Steady-state kinetics were established using AMV RT to determine substrate specificities towards canonical dNTPs (N = G, C, T, A). Overall we found evidence that RNA templates containing inosine are likely to incorporate dC > dT > > dA, where reactivity in the presence of dA was found to be pH dependent (process abolished at pH 7.3); and that the absence of the C2-exocyclic amine, as displayed with templates containing 8-oxoI, leads to increased selectivity towards incorporation of dA over dC. The data will be useful in assessing the impact that the presence of inosine and/or oxidatively generated lesions have on viral processes and adds to previous reports where I codes exclusively like G. Similar results were obtained upon comparison of AMV and MMLV RTs.

Fig 1. Diagram showing structures of the nucleobases of interest.

Horizontal black arrows denote the role of inosine in the de novo purine synthesis of nucleotides (two arrows between I and G represent intermediate xanthosine in the process). Dashed blue curved arrow represents the formation of I from A, catalyzed by adenosine deaminases acting on RNAs. Vertical arrows represent the structure of one of the oxidation products of each purine at the C8-position (left). Purine nucleotides can undergo an anti→syn conformational change upon C8-modification (right). The values shown for the oxidation potentials were taken from literature and the experimental setup is not the same for all [ref. [26] for G/A; [27] for I; and [28] for 8-oxoA/8-oxoI/8-oxoG].

Fig 2. RTn of duplexes 1:5–4:5.

(A) sequence and T_m analysis of all duplexes in PBS buffer (10 mM sodium phosphate, 1 mM NaCl, 5 mM MgCl₂, pH 7.5), recorded via CD; (B) reverse transcription using AMV RT and duplexes 1:5–4:5 in the presence of low [AMV RT]; and (D) higher [AMV RT] under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰). (C) represents steady-state kinetics for the processes of interest.

Fig 3

Thermal denaturation transitions of duplexes 1:6–4:6, 1:7–4:7, and 1:8–4:8 (A); and reactivity dNTP incorporation in the presence of AMV RT using duplexes 1:6–4:6 (B), 1:7–4:7 (C) and 1:8–4:8 (D) along with (E) their corresponding steady state kinetics for the most efficient cases. All experiments were carried out under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰).

Fig 4

Thermal denaturation transitions of duplexes 9:5–9:8 and H-bonding patterns expected from I, 8-oxoI, and 8-BrI (left); and RTn using AMV on duplexes 9:5–9:8 in the presence of canonical dNTPs, where M = equimolar mixture of all dNTPs, (right). Steady-state kinetics data is shown in the lower left corner. All experiments were carried out under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰).

Fig 5. Thermal denaturation transitions of duplexes 1:10–4:10 and 9:10 as well as 1:11–4:11 and 9:11 along with the corresponding RTn experiment using AMV RT.

The gel corresponding to 1:11–4:11 and 9:11 is included in the S9 File, where only the ratios in cDNA synthesis varied. ΔT_m corresponds to the difference between RNA:DNA hybrids containing 10 and 11. All RTn experiments were carried out under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰).

Fig 6. Reverse transcription using HIV RT with RNA:DNA duplexes 1:5–4:5 along with the % conversion.

All experiments were carried out under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰).

Fig 7. Reverse transcription using HIV RT with RNA:DNA duplexes 1:6–4:6 / 1:7–4:7 / 1:8–4:8, along with % conversions (shown underneath each gel).

All experiments were carried out under buffered conditions (5 mM Tris-acetate, 7.5 mM Potassium Acetate, 0.8 mM Magnesium Acetate (10 μM free Mg²⁺), 1 mM DTT, pH 8.3, 37 ⁰).

Fig 8

Wobble base pairs of G:U (A); of G:A displaying potential steric hinderance with the exocyclic amine, which is not present on I:A base pairs (B); base pairing between 8-oxoG:A, which places the exocyclic amine away from the H-bonding face (C); and possible base pair of 8-bromoinosine with A and C (D).