Intrinsic and regulated properties of minimally edited trypanosome mRNAs

Abstract

Most mitochondrial mRNAs in kinetoplastids require extensive uridine insertion/deletion editing to generate translatable open reading frames. Editing is specified by trans-acting gRNAs and involves a complex machinery including basal and accessory factors. Here, we utilize high-throughput sequencing to analyze editing progression in two minimally edited mRNAs that provide a simplified system due their requiring only two gRNAs each for complete editing. We show that CYb and MURF2 mRNAs exhibit barriers to editing progression that differ from those previously identified for pan-edited mRNAs, primarily at initial gRNA usage and gRNA exchange. We demonstrate that mis-edited junctions arise through multiple pathways including mis-alignment of cognate gRNA, incorrect and sometimes promiscuous gRNA utilization and inefficient gRNA anchoring. We then examined the roles of accessory factors RBP16 and MRP1/2 in maintaining edited CYb and MURF2 populations. RBP16 is essential for initiation of CYb and MURF2 editing, as well as MURF2 editing progression. In contrast, MRP1/2 stabilizes both edited mRNA populations, while further promoting progression of MURF2 mRNA editing. We also analyzed the effects of RNA Editing Substrate Binding Complex components, TbRGG2 and GAP1, and show that both proteins modestly impact progression of editing on minimally edited mRNAs, suggesting a novel function for GAP1.

Figure 1.

Sequence classes and IPSs in CYb and MURF2 editing. (A) The percentage of normalized pre-edited (Pre), fully edited (Full) and partially edited (Partial) sequences from n = 5 replicates are shown. Error bars represent one standard deviation. (B) Schematic (not to scale) of the mRNAs CYb, MURF2, RPS12 and ND7-5′ with associated gRNAs. The percentage of sequences containing fully edited sequence up to the end of the second gRNA are indicated. RPS12 and ND7 data are from (8). Dark blue, never edited region; light blue, region requiring further editing; green, region of fully edited sequence. (C) Locations of IPSs in CYb and MURF2 relative to edited mRNA sequence and known gRNAs. The sequences of the 5′ edited regions of both CYb and MURF2 are shown. Bars below the sequence represent gRNAs as published by Koslowsky et al. (11). Thicker black bars represent gRNA anchoring regions while hashed gray bars represent coverage range of identified gRNA families. Underscores (_) are used to clarify the editing sites to which the numbers above the lines correspond (CYb ES580 and MURF2 ES440). AuG start codon are also underlined. Pre-edited and fully edited sequences were excluded from IPS calculations. IPSs shown are present in at least 4 out of n = 5 replicates. Black diamonds represent IPSs present in all five replicates. Outlined diamonds represent IPSs present in four replicates. Small red u's denote uridines added to the sequence through editing, while large U’s denote uridines encoded by the mitochondrial genome. Green asterisks (*) denote encoded uridines that have been deleted. (D) Degree of pausing above threshold for identified IPSs. For each IPS, the total counts sharing the corresponding editing stop site were averaged and the fold count above the outlier threshold was calculated. Error bars represent one standard deviation.

Figure 2.

Sequence analysis of the most abundant junctions arising at IPSs in CYb mRNAs. (A) A family of junction sequences at Editing Stop Site (ESS)558 whose sequences suggest editing by an alternative gRNA. Above this sequence family, CYb pre-edited and fully edited mRNA sequences are aligned with an alternative gRNA, alt-gRNA-CYb, that could direct this editing pattern (top). Average normalized counts (Avg Norm) and lengths of junction sequences (JL) are indicated. Dashes (-) are used to align the non-U nucleotides for readability. AuG start codons (or their corresponding pre-edited positions) are underlined. (B) Most abundant junctions arising at ESS558 (>100 average normalized count) aside from those shown in (A). gRNA sequences as identified by Koslowsky et al. (11) are shown. Editing sites that match the canonical number of U’s are highlighted in yellow. Editing sites other than ES568 with 8 U’s are highlighted in green. (C) Most abundant junctions arising at ESS569 (>100 average normalized count). Editing sites that match the canonical number of U’s are highlighted in yellow.

Figure 3.

Sequence analysis of the most abundant junctions arising at IPSs in MURF2. (A) Most abundant junctions arising at ESS441 (>100 average normalized count). MURF2 pre-edited and fully edited sequence, along with the gRNA sequence identified by Koslowsky et al. (11) as directing this sequence are shown above. Average normalized counts (Avg Norm) and lengths of junction sequences (JL) are shown below. Editing sites that match the canonical number of u's are highlighted in yellow, while editing sites matching modification by alternative gRNAs, alt-MURF2-gRNA1 and alt-MURF2-gRNA2, are highlighted in blue. Editing sites that are canonically edited and whose editing can also be directed by the alternative gRNA (alt-MURF2-gRNA1) shown in (B) are highlighted in green. Dashes (-) are used to align the non-U nucleotides for readability. AuG start codons (or their corresponding pre-edited positions) are underlined. (B) Alternative gRNAs that could direct the editing of the most abundant junction at ESS441. (C) Most abundant junctions arising at ESS446 (>100 average normalized count).

Figure 4.

Effect of RBP16 and MRP1/2 depletion on editing and stability of CYb and MURF2 mRNAs. (A) The percentage of normalized pre-edited, partially edited and fully edited CYb sequences in uninduced controls (AvgUn) and MRP1/2 or RBP16 knockdown cells. Uninduced cells represent an average of all uninduced lines in this study (n = 8); RNAi induced MRP1/2 or RBP16 cells represent n = 2 for each cell line. Error bars represent one standard deviation. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 (Student's t-test). (B) Total CYb mRNA levels in RNAi induced cells relative to uninduced cells were determined by qRT-PCR (n = 6). (C and D) As above in (A and B) for MURF2 mRNA.

Figure 5.

EPSs in MURF2 mRNA resulting from RBP16 and MRP1/2 depletion. (A) MURF2 edited mRNA sequence with EPSs determined by comparison of induced knockdown cells (n = 2 for each cell line) to controls comprised of all uninduced cell lines used in this study (n = 8) after removal of pre-edited sequences and renormalization. Diamonds represent the locations of EPSs present in both replicates in each knockdown cell line. Lighter colored diamonds represent EPSs at which both uninduced and induced replicates have <500 average renormalized count. The position of gMURF2 is shown; thick region indicates anchor region. Underscore (_) clarifies the position of ES440. The AuG start codon is also underlined. (B) Average number of sequences at each EPS for MRP1/2 knockdowns (left) and RBP16 knockdowns (right), and their uninduced controls (AvgUn).

Figure 6.

Effect of TbRGG2 and GAP1 depletion on editing and stability of CYb and MURF2 mRNAs. (A) The percentage of normalized pre-edited, partially edited and fully edited CYb sequences in uninduced controls (AvgUn) and TbRGG2 or GAP1 knockdown cells. Uninduced cells represent an average of all uninduced lines in this study (n = 8); RNAi induced TbRGG2 or GAP1 cells represent n = 2 for each cell line. Error bars represent one standard deviation. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 (Student's t-test). (B) Total CYb mRNA levels in RNAi induced cells relative to uninduced cells were determined by qRT-PCR (n = 6). (C and D) As in (A and B) for MURF2 mRNA.

Figure 7.

EPSs in CYb mRNA resulting from TbRGG2 and GAP1 depletion. (A) CYb edited mRNA sequence with EPSs determined by comparison of induced knockdown cells (n = 2 for each cell line) to controls comprised of all uninduced cell lines used in this study (n = 8) after removal of pre-edited sequences and renormalization. Diamonds represent the locations of EPSs present in both replicates in each knockdown cell line. Lighter colored diamonds represent EPSs at which both uninduced and induced replicates have <500 average renormalized count. The positions of CYb gRNAs are shown; thick regions indicate anchor regions and hatched regions indicate differences in gRNA families (11). Underscore (_) clarifies the position of ES580. The position of the AuG start codon is also underlined. (B) Average number of sequences at each EPS for TbRGG2 knockdowns (left) and GAP1 knockdowns (right), and their uninduced controls (AvgUn).

Figure 8.

EPSs in MURF2 mRNA resulting from TbRGG2 and GAP1 depletion. (A) MURF2 edited mRNA sequence with EPSs determine by comparison of induced knockdown cells (n = 2 for each cell line) to controls comprised on all uninduced lines used in this study (n = 8) after of pre-edited sequences and renormalization. Diamonds represent the locations of EPSs present in both replicates in each knockdown cell line. Lighter colored diamonds represent EPSs at which both uninduced and induced replicates have <500 average renormalized counts. Underscore (_) clarifies the position of ES440. The AuG start codon is also underlined. (B) Average number of sequences at each EPS for TbRGG2 knockdowns (left) and GAP1 knockdowns (right), and their uninduced controls (AvgUn).