Multi-transcriptomics identifies targets of the endoribonuclease DNE1 and highlights its coordination with decapping

Abstract

Decapping is a crucial step in mRNA degradation in eucaryotes and requires the formation of a holoenzyme complex between the decapping enzyme DECAPPING 2 (DCP2) and the decapping enhancer DCP1. In Arabidopsis (Arabidopsis thaliana), DCP1-ASSOCIATED NYN ENDORIBONUCLEASE 1 (DNE1) is a direct protein partner of DCP1. The function of both DNE1 and decapping is necessary to maintain phyllotaxis, the regularity of organ emergence in the apex. In this study, we combined in vivo mRNA editing, RNA degradome sequencing, transcriptomics, and small RNA-omics to identify targets of DNE1 and study how DNE1 and DCP2 cooperate in controlling mRNA fate. Our data reveal that DNE1 mainly contacts and cleaves mRNAs in the coding sequence and has sequence cleavage preferences. DNE1 targets are also degraded through decapping, and both RNA degradation pathways influence the production of mRNA-derived small interfering RNAs. Finally, we detected mRNA features enriched in DNE1 targets including RNA G-quadruplexes and translated upstream open reading frames. Combining these four complementary high-throughput sequencing strategies greatly expands the range of DNE1 targets and allowed us to build a conceptual framework describing the influence of DNE1 and decapping on mRNA fate. These data will be crucial to unveil the specificity of DNE1 action and understand its importance for developmental patterning.

Figure 1.

In vivo editing using HyperTRIBE identifies mRNA in direct contact with DNE1. A) Immunoblot showing the protein accumulation in transgenic lines used for HyperTRIBE and expressing either the ADAR catalytic domain (ADAR) used as a control or protein fusions between DNE1 and ADAR. B) Venn diagram showing the overlap in loci edited by ADAR-DNE1 or ADAR-DNE1^D153N. Significant A-to-G edits were considered with adjPv < 0.01, Log2FC > 1, and a minimum of 10 reads. C) Distribution of edits by DNE1 and DNE1^D153N on mRNAs. D) Schemes showing the edits by ADAR-DNE1^D153N on several transcripts. UTR, untranslated region; CDS, coding sequence; mRNA, messenger RNA.

Figure 2.

Degradome analysis by GMUCT identifies two opposite trends on DNE1 targets upon mutation in DNE1. A) Venn diagram showing the output of a differential GMUCT analysis between dne1 xrn4 and xrn4 and displaying the overlap between loci showing increased (up) and decreased (down) 5′P fragments. B) Plots showing the repartition of decreased 5′P on two loci presenting only decreased 5′P in dne1 xrn4, data are expressed in rpm or Log2 fold-change between xrn4 dne1 and xrn4. C) Plots showing the repartition of 5′P on three loci presenting both decreased and increased 5′P in dne1 xrn4, data are expressed in rpm or Log2 fold-change between xrn4 dne1 and xrn4. Differential 5′P in B) and C) were considered with Log2FC ≥ 1 (colored in red) or Log2FC ≤ 1 (colored in blue) and Pv < 0.05 following the DEXseq analysis between dne1 xrn4 and xrn4. Data sets from the three biological replicates were pooled to generate the graphs presented in B) and C). D) Histogram showing the distribution on mRNAs of 5′P depending on their behavior in dne1 xrn4. E) Analysis of the nucleotide composition around the 1,296 main DNE1-dependent 5′P site using a sequence logo. The upper panel shows a control sequence logo produced using unchanged 5′P sites in dne1 xrn4 coming from the 1,296 loci producing DNE1-dependent 5′P. The lower panel shows the same analysis using the main DNE1-dependent 5′P from each locus. Position 0 represents the first nucleotide of the 5′P as sequenced in GMUCT. UTR, untranslated region; CDS, coding sequence; 5′P, 5′ monophosphate mRNA fragments; RPM, reads per million; log2FC, Log2 fold-change; no diff, not differential.

Figure 3.

Analysis of mRNA features enriched in mRNAs identified in HyperTRIBE and GMUCT. A) Venn diagram showing the overlap between loci edited by ADAR-DNE1^D153N and loci producing DNE1-dependent 5′P fragments. B) Venn diagram showing the overlap between loci edited by ADAR-DNE1^D153N and transcripts containing validated RNA G-quadruplex (rG4). C) Boxplot analysis of the number of introns and of mRNA, 5′ and 3′ UTR lengths for the DNE1-dependent loci identified by the different methods. Significantly different values (adjPv < 0.001) are labeled by different letters (Wilcoxon rank sum test). D) Proportion of transcripts containing uORFs or rG4 in the different lists of DNE-dependent loci based on refs. Significantly different values (adjPv < 0.001) are labeled by different letters (2-samples z-test of proportions). Boxplot displays the median, first and third quartiles (lower and upper hinges), the largest value within 1.5 times the interquartile range above the upper hinge (upper whisker), and the smallest value within 1.5 times the interquartile range below the lower hinge (lower whiskers). In C) and D), the lists of transcripts expressed in flowers and seedlings are used as control. Sample sizes are: control flowers n = 24,694, control seedlings n = 23,109, TRIBE D153N n = 2,252, GMUCT down n = 1,296, GMUCT Nagarajan n = 218, TRIBE D153N and GMUCT n = 288. UTR, untranslated region; CDS, coding sequence; uORF, upstream open reading frame; RNA G4, RNA G-quadruplex; Ribo-seq, ribosome profiling; rG4-seq, RNA G-quadruplex structure sequencing; nt, nucleotide; nb, number.

Figure 4.

Transcriptomic analysis of dcp2, dne1 dcp2, and xrn4 mutants identifies commonly deregulated transcripts. A) Plot showing the number of differentially expressed genes in dne1, dcp2, dne1 dcp2, and xrn4 versus WT with adjPv < 0.05 (n = 3). B) Venn diagram showing commonly upregulated loci between the two dne1 dcp2 double mutants and xrn4. C) Heatmap showing the mRNA accumulation pattern in dne1, dcp2, dne1 dcp2, and xrn4 for loci upregulated in both dne1 dcp2 double mutants. D) Predicted expression patterns of AT1G06150 (LHL1) and AT2G31280 (LHL2) in the shoot meristem of Arabidopsis using the 3D flower meristem tool from single cell experiments performed in Neumann et al. (2022). E) Venn diagram showing the overlap between upregulated loci in both dne1 dcp2 double mutants and loci identified by GMUCT and HyperTRIBE. WT, wild-type; vs, versus.

Figure 5.

Differential analysis of small RNA accumulation in dcp2, dne1 dcp2, and xrn4 mutants. A) Bar plots showing the output of the differential analysis of sRNA accumulation comparing mutants versus WT with adjPv < 0.05 (n = 3). B) Venn diagram showing the overlap observed for upregulated sRNAs between different mutants. C) Bar plots showing the output of the differential analysis of sRNA accumulation comparing dne1 dcp2 versus dcp2. D) Venn diagram showing the overlap observed for upregulated and downregulated sRNAs between the two dne1 dcp2 double mutants. E) RNA gel blot showing sRNA accumulation for loci differentially accumulating in dne1 dcp2 vs dcp2. Mean values, determined with ImageJ on blots from three biological replicates, are shown. The 21 nt size was determined by hybridization with an antisense probe targeting miR160. U6 was used as a loading control. F) Plots showing the accumulation of mRNA-derived siRNAs along the transcripts for loci with upregulated and downregulated siRNAs. Data sets from the three biological replicates were pooled to generate these graphs. G) Boxplot analysis of the number of introns and of mRNA, 5′ and 3′ UTR lengths for transcripts with differential sRNA accumulation in xrn4, dcp2, and dne1 dcp2. Significantly different values (adjPv < 0.001) are labeled by different letters (Wilcoxon rank sum test). H) Proportion of transcripts containing uORFs or rG4 in the different lists of transcripts with differential sRNA accumulation. Significantly different values (adjPv < 0.001) are labeled by different letters (2-samples z-test of proportions). Boxplot displays the median, first and third quartiles (lower and upper hinges), the largest value within 1.5 times the interquartile range above the upper hinge (upper whisker), and the smallest value within 1.5 times the interquartile range below the lower hinge (lower whiskers). In G) and H), the list of transcripts expressed in flowers is used as control. Sample sizes are: control flowers n = 24,694, Up dne1-2 dcp2 vs dcp2 n = 69, Up dne1-3 dcp2 vs dcp2 Up n = 67, Up dne1 dcp2 vs dcp2 n = 52, Up xrn4 n = 4,737, Up dcp2 n = 2,386, Down dne1-2 dcp2 vs dcp2 n = 123, Down dne1-3 dcp2 vs dcp2 n = 126, and Down dne1 dcp2 vs dcp2 n = 126. WT, wild-type; vs, versus; siRNA, small interfering RNA; nt, nucleotide; UTR, untranslated region; CDS, coding sequence; uORF, upstream open reading frame; RNA G4, RNA G-quadruplex; Ribo-seq, ribosome profiling; rG4-seq, RNA G-quadruplex structure sequencing; RPM, reads per million; nb, number.

Figure 6.

Diverse HTS techniques identify specific and common mRNAs influenced by DNE1. Bubble chart showing the extent of intersection between the list of loci identified by HyperTRIBE, GMUCT, sRNAseq, and RNAseq. Lists reported in the chart are as follows: for HyperTRIBE the results obtained with DNE1^D153N, for GMUCT loci with fewer 5′P in xrn4 dne1 vs xrn4, for siRNA and RNAseq lists obtained comparing dne1 dcp2 to dcp2. Each column corresponds to a list of loci, and each row corresponds to a possible intersection. Bubbles indicate the number of loci for each intersection with colors showing the number of related lists. siRNA, small interfering RNA; RNAseq, RNA sequencing.

Figure 7.

Models of DNE1 and DCP2 coordinated action on mRNAs. A, B) Integrated models for the action of DNE1 and DCP2 on mRNA-derived siRNA production. C) Integrated model built from the HyperTRIBE and GMUCT data. The model shows interaction and action of DNE1 in the CDS on sites with preferred nucleotide composition. Enriched features in DNE1 targets including RNA G4 and translated uORFs are depicted. WT, wild-type; siRNA, small interfering RNA; nt, nucleotide; UTR, untranslated region; CDS, coding sequence; uORF, upstream open reading frame; G4, RNA G-quadruplex; 5′P, 5′ monophosphate mRNA fragments.