Induro-tRNAseq
| Full name | Induro-tRNAseq |
| Summary | Using Induro a group-II intron-encoded RT to map and quantify genome-wide tRNA modifications 5 human cell lines and 3 mouse tissues |
| Key findings | In mouse cerebellum, neurodegeneration-associated tRNA-Arg-TCT-4 mutation C50U leads to lower tRNA expression, without changing the tRNA modifications. Mouse cerebellum shows lower acp3U20 modification than kidney and spleen, correlating with lower expression of enzyme DTWD1. tRNA isodecoders could have ~ 900-fold whereas tRNA isoacceptors could have ~ tenfold difference in expression across cell lines or tissue types |
| Sequencing strategy | RNA-seq |
| Raw data | https://www.ncbi.nlm.nih.gov/bioproject/996215 |
| Software/data repository | https://github.com/YaMing-Hou-lab/Induro-tRNAseq |
| Experimental protocol | Total RNA of cultured human cells was extracted with TRIzol (Invitrogen, #10296010). To determine tRNA charging levels, extracted total RNA was dissolved in 25 mM sodium acetate, pH 5.0. With mouse tissues, 12-week-old C57BL/6 J (B6J) WT mice were extracted for total RNA of cerebellum, kidney, and spleen, while the age-matched congenic B6J.B6NnTr20 mice were extracted for total RNA of cerebellum59. The mouse brain, kidney, and spleen used in this study were obtained from animals that were maintained under the institutional IACUC guidelines, compliant with those of the American Veterinary Medical Association for husbandry and euthanasia. The tRNA pool was isolated from total RNA by 12% denaturing PAGE/7 M urea in TBE (90 mM Tris-borate pH 8.3 and 2 mM EDTA) as gel slices, which were crushed with a disposable pestle in a tRNA extraction buffer (0.3 M Na Acetate (NaOAc), pH 5.0, 0.25% SDS, and 1 mM EDTA) as described18,71. The tRNA pool was filtered through a Costar Spin-X 8163 centrifuge tube filter (Corning #8163), precipitated with isopropanol in the presence of 20 µg glycogen (RNA grade; Thermo Scientific #R0551), and dissolved in RNase-free water. Approximately 12% of the input total RNA was recovered as tRNA. The tRNA in each sample was first deacylated by incubation for 45 min at 37 oC in 10 µL 75 mM glycine, pH 9.5, with 1 U/µL Superase-In (Thermo Fisher #AM2694). The reaction was diluted to 100 µL with 1X T4 PNK buffer and incubated for 30 min with 1 µL T4 PNK (10 U/µL; NEB M0201) to remove 3’ phosphoryl groups. The deacylated RNA was ethanol precipitated with 20 µg of glycogen and the pellet was dissolved in 7.0 µL water. To determine tRNA charging levels, 4 µL of 90 mM sodium acetate buffer, pH 4.8-5.0, was added to 500 ng of total tRNA (as part of ~2.5 µg total RNA) in 28 µL of water. Next, 4 µL of a freshly prepared 150 mM NaIO4 solution (Thermo Fisher #419610050) was added to give final concentrations of 16 mM NaIO4, 10 mM NaOAc, pH 4.8–5.0. The oxidation reaction was incubated 30 min at room temperature (r.t.) and was then quenched with the addition of 4 µL of 1 M glucose to 0.1 M final concentration and incubated 5 min at r.t. Next, 20 µL of 100 mM sodium borate, pH 9.5, was added for a final concentration of 33 mM. This mixture was incubated for 30 min at 45 °C for β-elimination and deacylation. The reaction was neutralized, and 3´-end repair was initiated by addition of 20 µL of 200 mM Tris-HCl, pH 6.8, 40 mM MgCl2, and 1 µL T4 PNK, and incubated at 37 °C for 30 min. The T4 PNK was heat-inactivated at 65 °C for 10 min. The deacylated RNA was ethanol precipitated with 20 µg of glycogen and the pellet was dissolved in 7.0 µL water. T4 Rnl2 (NEB M0239L) was used for splint-mediated 3’-ligation of each tRNA library to a barcoded 35-mer RNA/DNA adapter bearing a 5’-phosphate and a 3’-amino group (oligonucleotides listed in Tables S1, S2). Typically, 4 ligation reactions were carried out in parallel with each in 20 µL and containing a separate tRNA library, a uniquely barcoded 3’-adapter (0.25 µM), two paired DNA splints (0.32 µM) with both sharing the same barcode as the 3’-adapter, T4 Rnl2 (0.5 U/µL), and Superase-In RNase inhibitor (1 U/µL) in a buffer containing 400 µM ATP, 10 mM MgCl2, 1 mM DTT, 10% PEG8000, and 50 mM Tris-HCl, pH 7.5. The sequences of the two corresponding splint DNAs are listed in Table S2− one for the 3’-NCCA end derived from charged tRNAs and the other for the 3’-NCC end derived from uncharged tRNAs. Reactions were briefly heat-cooled at 60 °C for 2.5 min prior to adding T4 Rnl2 and kept at 16 °C for 45 min and then at 25 °C for 45 min. Ligation was terminated by combining all 4 reactions into 20 µL 0.1 M EDTA, followed by extraction with phenol-chloroform-isoamyl alcohol (25:24:1), pH 5.2, and purified through a Zymo Oligo Clean & Concentrator cartridge (Zymo Research #D4061). The nucleic acid in the multiplexed library was eluted in 11 µL of RNase-free water. Analysis of the tRNA ligated with the 3’-adapter was performed on a denaturing gel (Fig. 1b). Total RNA isolated from cells or a gel-purified tRNA pool was used as the input. The migration position of the starting tRNA was marked by a bracket (indicating a range of sizes), while the ligated tRNA was shown by an arrow. The intensity of each band was corrected by the molecular mass of each species and by removing the contaminating 5S rRNA band from the ligated band. The yield of ligation was measured by dividing the band intensity of the ligated tRNA (corrected for adapter addition and 5S RNA contamination) over the sum of the starting tRNA and the ligated tRNA, showing typically ≥ 75%. Each library was converted to cDNA with Induro RT (NEB M0681S) using a 57-mer ssDNA as the primer: 5’pRNAGATCGGAAGAGCGTCGTGTAGGGAAAGAG/iSp18/GTGACTGGAGTTCAGACGTGTGCTC-3’, where p is a phosphate, R is purine, N is any nucleotide, and iSP is an 18-carbon chain spacer. Each RT reaction was performed in 20 µL with 0.5 µM RT primer, 0.75 mM dNTPs, 10 mM DTT, 1 U/µL Superase-In, and 10 U/µL Induro RT in 1X Induro buffer. The reaction was briefly heat-cooled at 60 °C prior to adding the RT and was incubated overnight (12-16 h) at 42 °C unless otherwise specified. After cDNA synthesis, the RNA was hydrolyzed by addition of 1 µL of 5 M KOH and heated at 90-95 °C for 3 min. The solution was neutralized with 0.25 M NaOAc, pH 5.0, extracted with phenol-chloroform-isoamyl alcohol (25:24:1), and the cDNA precipitated with 3 volumes of ethanol. The pellet was dissolved in 8 µL of 7 M urea-TBE with 0.01% xylene cyanol and bromophenol blue. After heat-cool at 80-85 °C for 2 min, the cDNA was loaded onto a denaturing 10% PAGE/7 M urea gel (8 × 7 × 0.1 cm) and electrophoresed at 200 V in hot TBE for 20 min. Included in the gel as controls were an RT reaction conducted in the absence of tRNA and a Small Range RNA Ladder (NEB N0364S). The gel was stained with SYBR gold and visualized by a ChemiDoc imager (BioRad). The gel was excised in the range of 90–180 nucleotides (nts) to include full-length and truncated cDNA and was extracted into 500 µL TE by incubation in a MultiTherm Shaker (Benchmark Scientific) at 70 °C/1500 rpm for 1-2 h. After gel removal, the cDNA was precipitated with isopropanol in the presence of 20 µg glycogen carrier and the pellet dissolved in 5.5 µL water. The cDNA was circularized by incubation with Circligase (Lucigen CL4111K) for 3 h at 60 °C in a 10 µL reaction, which was terminated by heat inactivation (at 80 °C for 10 min) and stored at –20 °C until PCR amplification. The circularized cDNA (1 µL) was amplified by PCR using Q5 Hot Start High-Fidelity 2X Master Mix (NEB M0494L) in a 25 µL reaction containing 12.5 pmoles of forward and reverse primers (primers listed in Table S3). A unique barcoded reverse primer was used for each group of 4 multiplexed cDNA pools. The thermocycler was programmed for 30 sec at 98 °C, followed by at least 5 cycles of 10 sec at 98 °C, 10 sec at 62 °C, and 10 sec at 72 °C. To determine the optimal cycle number, separate reactions were terminated after 5, 6, 7, and 8 cycles. After clean-up by ethanol precipitation, these reactions were electrophoresed on a non-denaturing 8% PAGE gel alongside O’RangeRuler 10 bp DNA (Thermo Fisher #SM1313) and GeneRuler 50 bp DNA ladders (Thermo Fisher #SM0371), followed by SYBR gold staining. The dsDNA product (150-240 bp in length) with the highest yield but without contaminating higher MW DNA was excised from the gel and eluted into 500 µL TE by continuous mixing at r.t. overnight. After clarification of the suspension through a Costar Spin-X cartridge, the cDNA was ethanol precipitated with 20 µg glycogen and dissolved in 20 µL TE. The average yield of gel-purified cDNA from each multiplexed sample of 4 was approximately 50 ng as determined by Qubit (Invitrogen) and Bioanalyzer (Agilent) analysis. At least 20 ng of each multiplexed PCR product should be provided for sequencing. If the yield of PCR product was insufficient, the gel purification step could be eliminated, resulting in a 2-4-fold increase in the amount of recovered cDNA. In this case a Zymo Oligo Clean & Concentrator cartridge should be used to remove the PCR primers. Equimolar amounts of the PCR-amplified cDNA libraries were pooled and loaded onto the Illumina NextSeq 500 platform and a 2×75 paired-end sequencing run was performed. |
| Analysis protocol | aired-end reads were merged using PEAR v 0.9.6, which were then demultiplexed using cutadapt v2.5 and a fasta file of the first 10 nts for the different 3’-barcoded adapters was created. Indels in the alignment to the adapter were removed with --no-indels. The two 5’-RN nucleotides introduced by reverse transcription (see Library Preparation above) were trimmed from reads with -u 2. Reads shorter than 10 nts were discarded by cutadapt77 according to the parameters in -m 10. Processed reads of human samples were mapped to tRNA reference transcripts derived from human genome hg38 and those of mouse samples mapped to the reference derived from the mouse genome mm39 using mimseq v1.2. Mapping was as described (github.com/nedialkova-lab/mim-tRNAseq), where matured and processed tRNA sequences were mapped to MODOMICS entries using BLAST and clustered with the --cluster parameter using a user-defined sequence identity threshold. After clustering, reads were aligned using GSNAP to the representative cluster sequences of mature tRNAs. The following mapping parameters were used: H. sapiens: --species Hsap --cluster --cluster-id 0.95 --min-cov 2000 --max-mismatches 0.1 --remap --remap-mismatches 0.075 M. musculus: --species Mmus --remap --remap-mismatches 0.075 Mapping rates were calculated by the number of uniquely mapped reads relative to the sum of the number of unmapped reads, multiple-mapped reads, and uniquely mapped reads. Additional quantification and statistical analysis were also based on methods described34. Fractions of incomplete 3’-ends (3’-N, 3’-NC and 3’-NCC, where N = the discriminator nucleotide at position 73) relative to all 3’-ends (including 3’-NCCA) were calculated per unique tRNA sequence. For differential tRNA abundance analysis, read counts of cyto-tRNAs and mt-tRNAs were normalized with DESeq2 separately and the Pearson correlation coefficient (r) was calculated. Rates of misincorporation were calculated by summing up counts of mismatches for all four nucleotides relative to total read counts at the position of interest. Frequency of stops was determined by dividing the number of stops at a given position by the total number of reads at that position. Readthrough at a position was obtained by subtracting the ratio of RT stops at that position from 1.0. For a given modification, RT stops were reported as the maximum value observed in a 3-nt window (–1, 0, and +1) centering on the modification to reduce the likelihood of over-estimation. Modifications in human tRNAs at known positions (9, 20, 26, 32, 34, 37, and 58), as well as non-annotated positions, were identified by RT stop or RT misincorporation of >10%. All data analysis was performed with in-house python written codes using Python v3.7.0. D. melanogaster: --species Dmel --cluster --cluster-id 0.95 --min-cov 2000 --max-mismatches 0.1 --remap --remap-mismatches 0.075 |
| Organism/cell line | H. sapiens (HEK293T, K562, HeLa, SH-SY5Y, HAP1), M. musculus (brain, kidney, spleen) |
| Conditions | sodium arsenite stress |
| Approximate experimental time | 3-4 days |
| Starting RNA amount | 500 ng total RNA |
| tRNA expression | ++ |
| Base modifications | + |
| Charging status | - |
| tRNA processing and fragmentation | - |
| Citation | https://doi.org/10.1038/s41467-025-56348-1 |