Evidence for U-tail stabilization of gRNA/mRNA interactions in kinetoplastid RNA editing

Abstract

The most dramatic example of RNA editing is found in the mitochondria of trypanosomes. In these organisms, U-insertions/deletions can create mRNAs that are twice as large as the gene that encodes them. Guide RNAs (gRNAs) that are complementary to short stretches of the mature message direct the precise placements of the U residues. The binding of gRNA to mRNA is a fundamental step in RNA editing and understanding the relative importance of the elements that confer affinity and specificity on this interaction is critical to our understanding of the editing process. In this study, we have analyzed the relative binding affinities of two different gRNA/mRNA pairs. The affinity of gA6-14 for its message (ATPase 6) is high, with an apparent K(D) in the 5-10 nM range. In contrast, gCYb-558 has a low affinity for its cognate mRNA. Deletion of the gRNA U-tail caused a significant reduction in the binding affinity for only the gCYb-558 pair, and was observed only under physiological magnesium conditions. These results indicate that the U-tail contribution can differ substantially between the different gRNA/mRNA pairs. In addition, our results suggest that the efficiency of gRNA/mRNA interaction is highly dependent on thermodynamic parameters determined by the local sequences and their adopted structures surrounding the anchor-binding site.

Figure 1

Primary and secondary structure information for the gRNA/mRNA pairs. A. The primary sequence for the guide RNAs, gA6-14 and gCYb-558 and mRNAs 3’A6PES1END and 5’CYbUT. The anchor regions involved in the initial recognition between gRNA and mRNA are shown in bold and underlined. Sequence contributed by the vector is shown in lower case. B. Anchor Duplex interaction for gA6-14/A6PES1END and gCYb-558/5’CYbUT. Watson-Crick (|) and non-Watson-Crick (:) base pairs are indicated. C. Predicted secondary structures of A6PES1END and 5’CYbUT. Solution structure probing of A6PES1END indicates that it has very little structure. The structure shown is that predicted by mfold version 3.0 software [47]. The 5’ end of CYbUT forms a strong stemloop with the first few editing sites positioned within a terminal loop [11]. The nts of the mRNA anchor sequences are shown using the outline format. ΔGs were calculated using mfold version 3.0 [47].

Figure 1

Primary and secondary structure information for the gRNA/mRNA pairs. A. The primary sequence for the guide RNAs, gA6-14 and gCYb-558 and mRNAs 3’A6PES1END and 5’CYbUT. The anchor regions involved in the initial recognition between gRNA and mRNA are shown in bold and underlined. Sequence contributed by the vector is shown in lower case. B. Anchor Duplex interaction for gA6-14/A6PES1END and gCYb-558/5’CYbUT. Watson-Crick (|) and non-Watson-Crick (:) base pairs are indicated. C. Predicted secondary structures of A6PES1END and 5’CYbUT. Solution structure probing of A6PES1END indicates that it has very little structure. The structure shown is that predicted by mfold version 3.0 software [47]. The 5’ end of CYbUT forms a strong stemloop with the first few editing sites positioned within a terminal loop [11]. The nts of the mRNA anchor sequences are shown using the outline format. ΔGs were calculated using mfold version 3.0 [47].

Figure 2

Binding of gRNA to cognate pre-edited mRNA at varying concentrations of magnesium. Representative autoradiographs of 6% polyacrylamide gels, (A) 3’A6PES1END + gA6-14 and (B) 5’CYbUT + gCYb-558. Samples contained 5nM ³²P-labeled gRNA and increasing concentrations of cognate pre-edited mRNA. Odd number lanes - +U₁₅-tail gRNAs, Even number lanes – No U₁₅-tail gRNAs. The positions of the free gRNAs are indicated with bold arrows. indicate the positions of the major gRNA/mRNA complexes. For the gA6-14/3’A6PES1 pair a second complex (*) of slower mobility was observed at high mRNA concentrations. For the gA6-14/3’A6PES1 interaction, the following concentrations of pre-edited mRNAs were employed: 0 mRNA (lanes 1, 2), 1.25nM (lanes 3, 4), 2.5nM (lanes 5, 6), 5.0nM (lanes 7, 8), 12.5nM (lanes 9, 10), 25nM (lanes 11, 12), 50nM (lanes 13, 14), 100 nM (lanes 15,16). For the gCYb-558/CYbUT interaction, the following concentrations were used: 0 mRNA (lanes 1, 2), 0.5µM (lanes 3, 4), 0.75µM (lanes 5, 6), 1.0µM (lanes 7, 8), 1.5µM (lanes 9, 10), 2.0µM (lanes 11, 12), 2.25µM (lanes 13, 14), 2.5µM (lanes 15, 16).

Figure 3

Analysis of gA6-14/3’A6PES1END complexes at various Mg²⁺ concentrations. Quantification of complex formation was done using a Storm phosphorimager (Molecular Dynamics). The apparent affinity constant (K_D) was determined by fitting the data to the equation [complex]=[gRNA_f]×[mRNA]/([gRNA_f]+K_D) where [gRNA_f ] is the concentration of the free gRNA and [mRNA] is the concentration of unlabeled mRNA utilizing KaleidaGraph 3.5. The equation is based on the Langmuir binding isotherm that calculates the affinity constant based on the concentration of free gRNA that binds half of the maximum concentration of mRNA (16,18)[18]. Error bars indicate the standard deviation in complex formation observed. The calculated error in apparent K_D is shown in parentheses.

Figure 4

Analysis of gCYb-558/5’ CYbUT complexes at various Mg²⁺ concentrations and 2mM Ca²⁺. Apparent affinity constants (K_D) were calculated as explained in Figure 3. When the CYb gel shift data did not reach an experimental maximum binding, we used a derivation of the above equation (Lineweaver-Burk equation: 1/[complex] = 1/[gRNA_f] × K_D/[mRNA] + 1/[mRNA]) to calculate the K_D and B_LMAX (18).