Pentatricopeptide repeat poly(A) binding protein KPAF4 stabilizes mitochondrial mRNAs in Trypanosoma brucei

Abstract

In Trypanosoma brucei, most mitochondrial mRNAs undergo editing, and 3' adenylation and uridylation. The internal sequence changes and terminal extensions are coordinated: pre-editing addition of the short (A) tail protects the edited transcript against 3'-5' degradation, while post-editing A/U-tailing renders mRNA competent for translation. Participation of a poly(A) binding protein (PABP) in coupling of editing and 3' modification processes has been inferred, but its identity and mechanism of action remained elusive. We report identification of KPAF4, a pentatricopeptide repeat-containing PABP which sequesters the A-tail and impedes mRNA degradation. Conversely, KPAF4 inhibits uridylation of A-tailed transcripts and, therefore, premature A/U-tailing of partially-edited mRNAs. This quality check point likely prevents translation of incompletely edited mRNAs. We also find that RNA editing substrate binding complex (RESC) mediates the interaction between the 5' end-bound pyrophosphohydrolase MERS1 and 3' end-associated KPAF4 to enable mRNA circularization. This event appears to be critical for edited mRNA stability.

Fig. 1

Repeat organization, subcellular localization, and complex association of KPAF4. a Schematic repeat organization of kinetoplast polyadenylation factor 4 from Trypanosoma brucei (Tb) and Leishmania infantum (Li). Repeat boundaries were determined using the TPRpred online tool (https://toolkit.tuebingen.mpg.de/#/tools/tprpred) and adjusted according to Cheng et al. . Amino acids in positions 5 and 35/last potentially involved in adenosine recognition are indicated in separate columns. b Mitochondrial targeting of KPAF4-TAP fusion protein. Crude mitochondrial fraction was isolated by hypotonic lysis and differential centrifugation (crude mito), and further purified by renografin density gradient (pure mito). The latter preparation was extracted under conditions that separate matrix from membrane-bound proteins. Protein profiles were visualized by Sypro Ruby staining and KPAF4-TAP was detected with an antibody against the calmodulin-binding peptide. The mitochondrial enrichment was calculated by quantitative western blotting vs. total protein loading. Representative of two experiments is shown. c Tandem affinity purification of KPAF4. Final fraction was separated on 8–16% SDS gel and stained with Sypro Ruby. Representative of three experiments is shown. d KPAF4 co-purification with mRNA processing complexes. Fractions purified from parental cell line (beads, no tagged protein expressed), and mock and RNase-treated mitochondrial extracts were subjected to immunoblotting with antibodies against MERS1 NUDIX hydrolase (PPsome subunit), KPAP1 poly(A) polymerase, KPAF1 and KPAF3 polyadenylation factors, and GRBC1/2 (RNA editing substrate-binding complex, RESC) and RET1 TUTase (MPsome). Tagged KPAF4 was detected with antibody against calmodulin-binding peptide. RNA editing core complex (RECC) was detected by self-adenylation of REL1 and REL2 RNA ligases in the presence of [α-³²P]ATP. Representative of two experiments is shown. e Crude mitochondrial fraction was extracted with detergent and soluble contents were separated for 5 h at 178,000×g in a 10–30% glycerol gradient. Each fraction was resolved on 3–12% Bis–Tris native gel. Positions of native protein standards are denoted by arrows. KPAP1, KPAF4-TAP, MERS1, and GRBC1/2 were visualized by immunoblotting. REL1 and REL2 RNA ligases were detected by self-adenylation. Thyroglobulin (19S) and bacterial ribosomal subunits were used as apparent S-value standards. In each panel, representative of three experiments is shown

Fig. 2

KPAF4 interactions and proximity networks. a Model of the interactions between KPAF4, KPAP1 poly(A) polymerase, KPAF1-2, and KPAF3 polyadenylation factors, RNA editing substrate-binding complex (RESC), and MRP1/2 RNA chaperones. KPAP1, KPAF1, KPAF2, KPAF3, KPAF4, MRP2, MERS1, and Tb927.3.2670 proteins (encircled in blue) were affinity purified from mitochondrial lysates. The network was generated in Cytoscape software from bait–prey pairs in which the prey protein was identified by five or more unique peptides. The edge thickness correlates with normalized spectral abundance factor (NSAF) values ranging from 2.9 × 10⁻³ to 4.4 × 10⁻⁵ (Supplementary Data 1). Edges between tightly bound RESC modules (GRBC, REMC, and PAMC) were omitted for clarity. All purifications were performed in parallel under uniform conditions. b KPAF4 proximity network. Spectral counts derived from BioID experiments with KPAP1, KPAF4, GRBC2, and MERS1 fusions with BirA* biotin ligase (encircled in blue) were processed as in (a) to build a proximity network. The edge thickness correlates with normalized spectral abundance factor (NSAF) values ranging from 2.9 × 10⁻³ to 2.6 × 10⁻⁵ (Supplementary Data 2). Edge colors other than red are for visualization purposes only. All purifications were performed in parallel under uniform conditions

Fig. 3

KPAF4 repression effects on cell growth and polyadenylation complex. a Northern blotting analysis of KPAF4 mRNA downregulation by inducible RNAi. b Growth kinetics of procyclic parasite cultures after mock treatment and KPAF4 RNAi induction with tetracycline. Data representative of three independent experiments are shown as mean ± s.d. c Quantitative real-time RT-PCR analysis of RNAi-targeted KPAF4 mRNA, and mitochondrial rRNAs and mRNAs. The assay distinguishes edited and corresponding pre-edited transcripts, and unedited mRNAs. RNA levels were normalized to β-tubulin mRNA. RNAi was induced for 55 h. Error bars represent the standard deviation from at least three biological replicates. The thick line at “1” reflects no change in relative abundance; bars above or below represent an increase or decrease, respectively. P, pre-edited mRNA; E, edited mRNA. d Cell lysates prepared at indicated time points of KPAF4 RNAi induction were sequentially probed and re-probed by quantitative immunoblotting on the same membrane in the following order: antigen-purified rabbit polyclonal antibodies against KPAP1, KPAF1, KPAF3, GRBC1/2, and mouse monoclonal antibodies against RET1 TUTase. Signals were normalized against three independent loading standards and mean values calculated

Fig. 4

Divergent effects of KPAF4 knockdown on mitochondrial RNAs. a Northern blotting of pre-edited (Pre-E), partially edited (Part-E), and fully edited RPS12 mRNA variants. Total RNA was separated on a 5% polyacrylamide/8 M urea gel and sequentially hybridized with radiolabeled single-stranded DNA probes. Zero-time point: mock-induced RNAi cell line. Cytosolic 5.8S rRNA was used as loading control. Parent, RNA from parental 29-13 cell line; (dT), RNA was hybridized with 20-mer oligo(dT) and treated with RNase H to show positions of non-adenylated molecules in parental cell line. Pre-edited RNA length increase in KPAF4 RNAi is shown by brackets. Representative of four experiments for edited and three experiments for pre-edited RPS12 mRNA forms are shown. b Alignment of representative RPS12 mRNA 3′ ends in KPAF4 RNAi cells. RNA termini were amplified by cRT-PCR, cloned and sequenced. A fragment of 3′ untranslated region, short A-tail, and U-extensions are indicated. c Northern blotting of pan-edited A6 mRNA. Total RNA was separated on a 1.7% agarose/formaldehyde gel and sequentially hybridized with oligonucleotide probes for pre-edited and fully edited sequences. Loading control: cytosolic 18S rRNA. Representative of three experiments is shown. d Northern blotting of moderately edited cyb mRNA. Total RNA was separated on a 1.7% agarose/formaldehyde gel and hybridized with oligonucleotide probes for pre-edited and fully edited sequences. Loading control: cytosolic 18S rRNA. Representative of two experiments is shown. e Northern blotting of unedited CO1 and ND1 mRNAs. Total RNA was separated on a 1.7% agarose/formaldehyde gel and sequentially hybridized with oligonucleotide probes. Loading control: cytosolic 18S rRNA. Representative of two experiments is shown. f Northern blotting of mitochondrial ribosomal RNAs. Total RNA was separated on a 5% polyacrylamide/8 M urea gels and hybridized with oligonucleotide probes. Loading control: cytosolic 5.8S rRNA. Representative of two experiments is shown. g Guide RNA northern blotting. Total RNA was separated on a 10% polyacrylamide/8 M urea gel and hybridized with oligonucleotide probes specific for gA6(14) and gCO3(147). Mitochondrially localized tRNA^Cys served as loading control. Single experiment performed

Fig. 5

Sequencing of mRNA and rRNA 3′ extensions in KPAF4 RNAi background. a Length distribution of short mRNA and 12S rRNA tails. Non-encoded 3′ end extensions (MiSeq instrument, Illumina, single biological replicate) were individually binned into 10-nt length groups. Mock-induced and RNAi datasets, indicated by blue and red bars, respectively, represent percentage of the total number of reads. b Length distribution of long mRNA tails. Non-encoded 3′ end extensions (PacBio RS II instrument, two biological replicates) were individually binned into 10-nt length groups before 100 nt, and in 50-nt groups thereafter. Mock-induced and RNAi datasets are indicated by blue and red bars, respectively, that represent percentage of the total number of reads. c Positional nucleotide frequencies in short mRNAs and 12S rRNA tails. A nucleotide percentage was calculated for each position that contained at least 5% of the total extracted sequences. The nucleotide bases are color-coded as indicated. Arrows show positions of equal adenosine and uridine frequencies. d Positional nucleotide frequencies in long mRNA tails. A nucleotide percentage was calculated for each position that contained at least 5% of the total extracted sequences. The nucleotide bases are color-coded as indicated. Arrows show positions of equal adenosine and uridine frequencies

Fig. 6

Distribution of KPAF4 in vivo-binding sites between pre-edited and edited mRNAs. a Isolation of in vivo KPAF4-RNA crosslinks. Modified TAP-tagged fusion protein was purified by tandem affinity pulldown from UV-irradiated (+) or mock-treated (−) parasites. The second purification step was performed under fully denaturing conditions and resultant fractions were subjected to partial on-beads RNase I digestion and radiolabeling. Upon separation on SDS–PAGE, RNA–protein crosslinks were transferred onto nitrocellulose membrane. Protein patterns were visualized by Sypro Ruby staining (left panel), and RNA–protein crosslinks were detected by exposure to phosphor storage screen (right panel). RNA from areas indicated by brackets was sequenced. Representative of six biological replicates is shown. b KPAF4 in vivo-binding sites. Crosslinked fragments were mapped to the maxicircle’s gene-containing region. Annotated mitochondrial transcripts encoded on major and minor strands are indicated by blue and red arrows, respectively. c Position-specific nucleotide frequency in partially mapped KPAF4 CLAP-Seq reads. In reads selected by partial mapping to maxicircle and edited mRNAs, the unmapped 3′ segments were considered as tail sequences. The nucleotide frequency was calculated for each position beginning from the 3′ end. d Aggregate KPAF4 mRNA-binding pattern. Read coverage is represented by the gray area, and the nucleotides in 3′ extensions are color-coded at their projected positions. e KPAF4 binding to representative pan-edited (RPS12, A6) and moderately edited (CYB) mRNAs. Read coverage profiles were created for matching pre-edited and fully edited mRNA. Read coverage is represented by the gray area, and the unmapped nucleotides in 3′ extensions are color-coded at their projected positions. The mRNA is highlighted with a rose bar in the context of adjacent maxicircle sequences. f MERS1 binding to representative pan-edited (RPS12, A6), and moderately edited (CYB) mRNAs. Graphs were created as in panel e

Fig. 7

KPAF4-bound adenylated RNA is partially resistant to uridylation and degradation in vitro. a Western blotting of affinity purified KPAF4-WT and KPAF4-Mut samples. Protein samples were purified from mitochondrial fraction by rapid affinity pulldown with IgG-coated magnetic beads. KPAF4 polypeptides were detected with an antibody against the calmodulin-binding peptide. Single experiment performed. b Electrophoretic mobility shift assay with KPAF4-WT. Increasing amounts of affinity-purified KPAF4 were incubated with 5′ radiolabeled RNAs and separated on 3–12% native PAGE. Representative of six experiments is shown. c Electrophoretic mobility shift assay with KPAF4-Mut was performed as in (b). Representative of two experiments is shown. d RNA adenylation and uridylation. KPAP1, RET1, or in combination, were incubated with 5′ radiolabeled RNA and ATP, UTP, or ATP/UTP mix, respectively, in the absence or presence of KPAF4. Recombinant enzymes were purified from bacteria as described^,. Reactions were terminated at indicated time intervals and products were resolved on 10% polyacrylamide/ 8 M urea gel. Representative of seven experiments are shown. e RNA degradation. The same RNA substrates as in (d) were incubated with increasing concentrations of KPAF4, and the reactions were initiated by adding buffer or the MPsome for a fixed period of time (DSS1, left panels). RNAs were incubated with a fixed concentration of KPAF4 for 20 min, and reactions were initiated by adding the MPsome (right panel). Reactions were terminated at indicated time intervals and products were resolved on a 10% polyacrylamide/8 M urea gel. Input RNA and final degradation products of 4–5 nt (FP) are shown. Representative of two experiments are shown and quantified in Supplementary Fig. 4