Structure-function analysis of Methanobacterium thermoautotrophicum RNA ligase - engineering a thermostable ATP independent enzyme

Abstract

Background: RNA ligases are essential reagents for many methods in molecular biology including NextGen RNA sequencing. To prevent ligation of RNA to itself, ATP independent mutant ligases, defective in self-adenylation, are often used in combination with activated pre-adenylated linkers. It is important that these ligases not have de-adenylation activity, which can result in activation of RNA and formation of background ligation products. An additional useful feature is for the ligase to be active at elevated temperatures. This has the advantage or reducing preferences caused by structures of single-stranded substrates and linkers.

Results: To create an RNA ligase with these desirable properties we performed mutational analysis of the archaeal thermophilic RNA ligase from Methanobacterium thermoautotrophicum. We identified amino acids essential for ATP binding and reactivity but dispensable for phosphodiester bond formation with 5' pre-adenylated donor substrate. The motif V lysine mutant (K246A) showed reduced activity in the first two steps of ligation reaction. The mutant has full ligation activity with pre-adenylated substrates but retained the undesirable activity of deadenylation, which is the reverse of step 2 adenylation. A second mutant, an alanine substitution for the catalytic lysine in motif I (K97A) abolished activity in the first two steps of the ligation reaction, but preserved wild type ligation activity in step 3. The activity of the K97A mutant is similar with either pre-adenylated RNA or single-stranded DNA (ssDNA) as donor substrates but we observed two-fold preference for RNA as an acceptor substrate compared to ssDNA with an identical sequence. In contrast, truncated T4 RNA ligase 2, the commercial enzyme used in these applications, is significantly more active using pre-adenylated RNA as a donor compared to pre-adenylated ssDNA. However, the T4 RNA ligases are ineffective in ligating ssDNA acceptors.

Conclusions: Mutational analysis of the heat stable RNA ligase from Methanobacterium thermoautotrophicum resulted in the creation of an ATP independent ligase. The K97A mutant is defective in the first two steps of ligation but retains full activity in ligation of either RNA or ssDNA to a pre-adenylated linker. The ability of the ligase to function at 65°C should reduce the constraints of RNA secondary structure in RNA ligation experiments.

Figure 1

(A) Structural comparison of T4 RNA ligase 2 and archaeal RNA ligase PAB1020 active sites. Two structures were superimposed based on coordinates of pre-bound AMP and ATP homolog (AMPPNP) as well as known conserved amino acids of the ligase active sites. The T4Rnl2 structure is represented in grey and PAB1020 in yellow. The numbers of amino acids in the conserved motifs I-V (in parenthesis) are for T4Rnl2 and PAB1020 ortholog MthRnl, which were determined after sequence alignment of two archaeal enzymes. (B) The sequences of the conserved motifs I and V of MthRnl compared to corresponding motifs in the RNA and DNA ligases as discussed in the text. The listed RNA ligases are from: MthRnl (Methanobacterium thermoautotrophicum), PAB1020 (Pyrococcus abyssi), TS2126 (bacteriophage Thermus scotoductus), RM378 (bacteriophage Rhodothermus marinus), T4Rnl1 and T4Rnl2 (bacteriophage T4), AcNPV (Autographa californica nucleopolyhedrovirus). And DNA ligases are from: PBCV1 (Chlorella virus), MthDnl (Methanobacterium thermoautotrophicum), VacDnl (Vaccinia virus). The conserved lysines are shown in bold.

Figure 2

Activity of the MthRnl and the K246A, K246Q and K97A mutants in different steps of the ligation reaction.(A) pH-dependence of self-adenylation activity of enzymes (step 1) in reactions with radioactive ATP. The reactions were carried out as described in the Methods and the products were resolved by an SDS-PAGE. Autoradiographs of the gels are shown. (B) pH-dependence of adenylation of 5’-phosphorylated, 3’-amino-blocked ssDNA (pDNA17-NH₂) (step 2) with wild type and the mutants of MthRnl under reaction conditions described in the Methods. The reaction products were separated by 15% urea-PAGE and visualized with SYBR Gold stain. (C) pH-dependence of deadenylation/hydrolysis of 5’ pre-adenylated DNA (AppDNA17-NH₂) with wild type and the mutants of MthRnl under reaction conditions described in the Methods. The products were analyzed on 15% urea-PAGE and stained with SYBR Gold. Enzymes used in each step of ligation reactions are indicated on the left. Position of DNA donor substrates and products are indicated on the right. The pH of the buffer is indicated at the bottom. The control (c) indicates the reaction without enzyme.

Figure 3

ATP independence of the K97A MthRnl mutant.(A) The reactions of 5’ pre-adenylated 3’-amino-blocked DNA donor (AppDNA17-NH₂) with ssDNA acceptor (DNA25) and K97A MthRnl mutant without ATP or with increasing (5–500 μM) concentration of ATP. (B) The reaction of 5’-phosphorylated 3’-amino-blocked DNA donor (pDNA17-NH₂) with ssDNA acceptor (DNA25) and K97A MthRnl mutant without ATP or with increasing (5–500 μM) concentration of ATP. (C) The reactions of 5’ pre-adenylated 3’-amino blocked DNA donor (AppDNA17-NH₂) with ssDNA acceptor (DNA25) and wild type MthRnl without ATP or with increasing (5–500 μM) concentration of ATP. The ligation reactions were carried out as described in the Methods. The products were separated on 15% urea-PAGE and visualized with SYBR Gold stain. The concentration of ATP used in reactions is indicated at the bottom. Positions of substrates and ligated products are indicated on the right. Mr are molecular weight RNA markers. The ‘input’ indicates reactions without enzyme.

Figure 4

A comparison of the wild type MthRnl and the MthRnl mutants in step 3 of the ligation reaction with various acceptors. Enzymes were assayed using three different acceptors with identical sequence: RNA (RNA30), 2’-O-methylated at 3’-end RNA (RNA30-OMe) or ssDNA (DNA30), and the donor AppDNA17-NH₂ as described in the Methods. (A) The control reactions without enzyme. (B) The reactions with non-adenylated form of the wild type MthRnl. (C, D and E) The reactions with K97A, K246A and K246Q MthRnl mutants respectively. The products were separated on 15% urea-PAGE and visualized with SYBR Gold stain. Positions of substrates, ligated products and by-product pDNA17-NH₂ are indicated on the right. Acceptor substrates used in the reactions are indicated on the top. Mr indicates molecular weight RNA markers.

Figure 5

A comparison of the K97A MthRnl mutant activity with activities of truncated T4Rnl2 and T4Rnl1 in step 3 of the ligation reaction. Enzymes were assayed using three different acceptors with identical sequence: RNA (RNA30), 2’-O-methylated at 3’-end RNA (RNA-OMe) or ssDNA (DNA30) in reactions with 5’ pre-adenylated DNA donors AppDNA17-NH₂ or AppDNA21-NH₂ as described in the Methods. (A) Control reactions without enzyme. (B, C and D) Reactions with truncated T4Rnl2, K97A MthRnl mutant and T4Rnl1 respectively. (E) For comparison, T4Rnl1 was assayed in the ligation reactions with 5’-phosphorylated 3’-amino-blocked DNA donor (pDNA17-NH₂) and indicated acceptors in presence of ATP. (F) Percentage of ligation of the different acceptors was calculated from experiments shown in Figures 4 (panels B and C) and 5 (panels B, C and D). Mean values and standard deviation were calculated from multiple (2–5) gels analysis described in the Methods are shown.

Figure 6

The comparison of pre-adenylated RNA or DNA donors in step 3 of the ligation reaction with the K97A MthRnl mutant or truncated T4Rnl2. Enzymes were assayed using 2’-O-methylated at 3’-end RNA (RNA-OMe) acceptor in reactions with 5’ pre-adenylated RNA (AppRNA21-NH₂) or DNA (AppDNA21-NH₂) with identical sequence. (A) Reactions without (input) or with truncated T4Rnl2. (B) Reactions without (input) or with K97A MthRnl mutant. Reactions were carried out as described in the Methods at 25°C for T4Rnl2tr and 65°C for K97A mutant. The products were analysed on 15% urea-PAGE and stained with SYBR Gold stain. Positions of substrates and ligated products are indicated on the right. Donor substrates used in the reactions are indicated on the top. Mr indicates molecular weight RNA markers.