Human NSCLC cell lines H1299 and A549 were obtained from the Cell Bank of the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). All the cell lines were validated by STR profiling and tested negative for mycoplasma.

pCDH-Puro-3Flag was utilized as the vector to construct the ALKBH4 expression plasmid. Briefly, the ALKBH4 coding sequence was amplified using primers designed with homologous arms flanking insertion sites. Concurrently, the pCDH-Puro-3Flag vector was linearized through PCR amplification to eliminate the original coding sequence. Both PCR products were then subjected to in vitro recombination using the NovoRec^® One-Step PCR Cloning Kit (Novoprotein, Suzhou, China) in a molar ratio-based reaction system. The ligation product was transformed into TransStbl3 competent cells and plated onto LB agar plates supplemented with ampicillin for colony growth. Positive clones were screened utilizing colony PCR. The ALKBH4 expression plasmid was subsequently verified by sequencing (Sangon Biotech, Shanghai, China).

For H₂O₂ treatment, H₂O₂ solution (Sigma-Aldrich, MO, USA) was diluted in complete medium and cells were treated with H₂O₂ solution at designated concentrations for 24 h. The cells were then collected and analyzed. For plasmid transfection, cells were transiently transfected with 2 µg of ALKBH4 plasmid or pCDH-Puro-3Flag plasmid using Neofect DNA Transfection Reagent (Neofect, Beijing, China) according to the manufacturer’s instructions. For RNA transfection, cells were transiently transfected with mimics, antisense oligonucleotides (ASO), or corresponding controls at a final concentration of 50 nM using Lipofectamine^TM RNAiMAX reagent (Thermo Fisher Scientific, MA, USA) according to the manufacturer’s instructions. Cells were collected and analyzed 24 h after transfection. 5^′-tsRNA^Glu mimics and 5^′-tsRNA^Glu-targeting ASO were chemically synthesized (Sangon Biotech, Shanghai, China). The oligonucleotide sequences are listed in Supplementary Table 1.

For cell proliferation assays, cells were seeded in 96-well plates at a density of 1000 cells per well. Cell proliferation was subsequently measured using a Cell Counting Kit-8 (Bimake, TX, USA) according to the manufacturer’s instructions. The absorbance of each well was measured at 450 nm at each time point. For colony formation assays, cells were seeded in 6-well plates at a density of 1000 cells per well. After incubation at 37 °C for approximately 14 days, cells were washed three times with PBS, fixed with 4% paraformaldehyde for 30 min, and stained with 0.2% crystal violet solution for 30 min. After washing three times with PBS, each well was photographed, and the number of colonies was counted. Three independent experiments were performed, each in triplicate.

Cells were harvested and washed twice in PBS, suspended in 700 µL of 75% ethanol, mixed with 300 µL of PBS, and refrigerated at –20 °C overnight. Following centrifugation and washing twice with PBS, cells were resuspended, stained with propidium iodide (PI)/RNase Stain Binding Buffer, and placed at 4 °C for 30 min. DNA content was then measured using a FACSCalibur flow cytometer (BD Biosciences, CA, USA). Three independent experiments were performed, each in triplicate.

Experiments were performed using the Click-iT Plus OPP Protein Synthesis Assay Kit (Invitrogen, CA, USA). Briefly, cells in 6-well plates were incubated with 1 mL of medium containing 10 µM Click-iT Plus OPP working solution for 15 min. After centrifugation and washing with PBS, cells were fixed in 3.7% paraformaldehyde for 15 min, permeabilized in 0.5% Triton X-100 for 15 min, and then incubated in the dark with 100 µL of Click-iT Plus OPP reaction cocktail for 30 min. Subsequently, the cells were washed with the Click-iT reaction rinse buffer and stained with 4^′,6-diamidino-2-phenylindole (DAPI). Samples were then analyzed using a FACSCanto II flow cytometer (BD Biosciences, CA, USA). Three independent experiments were performed, each in triplicate.

Cells were treated with 100 µg/mL cycloheximide (CHX) for 5 min, lysed with lysis buffer (20 mM Tris-HCl, pH 7.5, 2.5 mM MgCl₂, 150 mM NaCl, 100 µg/mL of CHX, 1 mM DTT, 100 U/mL of RNasein, 0.5% Triton X-100) on ice; subsequently, the cells were centrifuged and the supernatant was collected. The supernatant was layered on top of a 5–45% (w/v) sucrose gradient and centrifuged at 32,000 rpm at 4 °C for 1.5 h. Subsequently, the sample was fractionated using a Gradient Fractionator (Biocomp, Quebec, Canada) while continuously monitoring the absorbance at 260 nm. Fractions were categorized and used to isolate total RNA for subsequent RNA analysis. Three independent experiments were performed, each in triplicate.

The 5^′-tsRNA^Glu target sequence (reverse complement to the 5^′-tsRNA^Glu sequence) was subcloned into the pmirGLO vector (Tsingke, Beijing, China). Subsequently, the cells were co-transfected with the reporter plasmid and 5^′-tsRNA^Glu mimics or NC mimics using Lipofectamin^TM 3000 (Invitrogen, CA, USA). 24 h post-transfection, firefly and renilla luciferase activities were quantified using the Dual-Luciferase Reporter Assay System (Beyotime, Shanghai, China) following the manufacturer’s protocol. Three independent experiments were performed, each in triplicate.

Cells were washed with PBS and harvested by centrifugation. Nuclear and cytoplasmic extracts were then prepared using a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, Shanghai, China) following the manufacturer’s instructions. Briefly, cell pellets were resuspended in Reagent A, vortexed, and then incubated on ice. Following the addition of Reagent B and vortexing, the mixture was incubated on ice and centrifuged to separate the supernatant (containing cytoplasmic proteins) from the pellet (containing nuclear proteins). Cytoplasmic and nuclear RNAs were extracted from these fractions by adding TRIzol^TM reagent (Invitrogen, CA, USA). Three independent experiments were performed, each in triplicate.

Biotin-labeled 5^′-tsRNA^Glu and its antisense probe were synthesized by Sangon Biotechnology (Shanghai, China). RNA pull-down assays were then conducted using the Pierce^TM Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific, MA, USA) according to the manufacturer’s instructions. Briefly, 100 pmol of biotin-labeled 5^′-tsRNA^Glu or its antisense probe was added to 50 µL of streptavidin-coated magnetic beads and incubated for 30 min under constant agitation. The Master Mix containing pre-prepared cell lysates was then added to the RNA-bound beads and incubated at 4 °C for 60 min. Subsequently, the beads were washed three times with wash buffer, resuspended in loading buffer, followed by denaturation at 100 °C for 10 min. Eluted proteins were then separated by SDS-PAGE and visualized using Coomassie Blue staining. Protein bands within the molecular weight range of 20–150 kDa were excised and subjected to protein identification using LC-MS/MS on a Q-Exactive HF X tandem mass spectrometer (Thermo Fisher Scientific, CA, USA) at BGI (Shenzhen, China). The sequences of all RNA pull-down probes are provided in Supplementary Table 1. Three independent experiments of RNA pull-down assay were performed, each in triplicate.

RIP assays were conducted using an Immunoprecipitation Kit with Protein A+G Magnetic Beads (Beyotime, Shanghai, China), according to the manufacturer’s protocol. Briefly, cells were lysed in RIP lysis buffer containing protease inhibitors and RNase inhibitors. The lysate was centrifuged, and then the supernatant was collected and incubated with 0.5  µg of RPL17 antibody (14121-1-AP, Proteintech, IL, USA) or IgG antibody (A7016, Beyotime, Shanghai, China) under rotation at 4 °C for 24 h. Subsequently, 10 µL of Protein A+G magnetic beads were added and incubated with rotation at 4 °C for 16 h. RNA bound to the immunoprecipitated complexes was extracted using TRIzol^TM reagent (Invitrogen, CA, USA) and subjected to quantitative RT-PCR analysis. Three independent experiments were performed, each in triplicate.

Total RNA was isolated from cell lines using TRIzol^TM reagent (Invitrogen, CA, USA), according to the manufacturer’s protocol. To detect small RNAs, cDNA was reverse-transcribed from 500 ng of total RNA using the PrimeScrip^TM Ⅱ 1st Strand cDNA Synthesis Kit (Takara, Tokyo, Japan). For mRNA detection, cDNA was synthesized from 500 ng of total RNA using a PrimeScript^TM RT Master Mix Kit (Takara, Tokyo, Japan). Quantitative real-time PCR was performed using 2$\times$SYBR Green qPCR Master Mix (Bimake, TX, USA). U6 and GAPDH served as internal controls to normalize the expression levels of 5^′-tsRNA^Glu and mRNAs, respectively. Relative quantification of targets was calculated using the 2^{-Δ⁢Δ⁢Ct} method. All primer sequences are provided in Supplementary Table 1.

Total RNAs were resolved on a 10% urea-PAGE gel and stained with SYBR Gold (Thermo Fisher Scientific, MA, USA), followed by immediate imaging. The separated RNAs were then transferred onto positively charged nylon membranes (Roche, Basel, Switzerland) and ultraviolet-crosslinked at an energy of 0.12 J. Subsequently, membranes were pre-hybridized with DIG Easy Hyb solution (Roche, Basel, Switzerland) for 1 h at 42 °C. After pre-hybridization, membranes were hybridized overnight at 42 °C with a DIG-labeled 5^′-tsRNA^Glu probe (Sangon Biotech, Shanghai, China). The probe sequences are listed in Supplementary Table 1. Subsequently, membranes were washed, blocked, and incubated with an anti-DIG-AP antibody (11093274910, 1:10,000, Roche, Basel, Switzerland) for 30 min. Signal detection was performed using the chemiluminescent substrate CSPD (Roche, Basel, Switzerland), and visualization was performed using a ChemiDoc^TM MP Imaging System (Bio-Rad, CA, USA).

Western blot analysis was performed according to a previously established protocol [40]. The following primary antibodies were used: rabbit monoclonal anti-ALKBH4 (ab195379, 1:1000, Abcam, Cambridge, UK), rabbit polyclonal anti-Cyclin D1 (AF0126, 1:1000, Beyotime, Shanghai, China), rabbit polyclonal anti-RPL17 (14121-1-AP, 1:1000, Proteintech, IL, USA), rabbit polyclonal anti-Cyclin B1 (55004-1-AP, 1:1000, Proteintech, IL, USA), rabbit polyclonal anti-Phospho-Histone-H3 (9701, 1:1000, Cell Signaling Technology, MA, USA), rabbit monoclonal anti-Histone-H3 (4499T, 1:1000, Cell Signaling Technology, MA, USA), rabbit polyclonal anti-GAPDH (AB-P-R001, 1:1000, Goodhere Biotech, Hangzhou, China), and rabbit monoclonal anti-$\beta$-actin (T40104, 1:1000, Abmart, Shanghai, China).

H1299 cells were transiently transfected for 24 h with the pCDH plasmid, ALKBH4 expression plasmid, NC mimics, and 5^′-tsRNA^Glu mimics, as indicated. Total RNA was extracted using TRIzol^TM reagent (Invitrogen, CA, USA). Following quantitative analysis and quality inspection, the RNA samples were subjected to sequencing by BGI (Shenzhen, China). Libraries were constructed using the Optimal Dual-mode mRNA Library Prep Kit (BGI, Shenzhen, China), and sequenced on a BGIseq500 platform (BGI, Shenzhen, China). Raw sequencing data were normalized for subsequent analysis. RNA-seq data are shown in Supplementary Table 2. Differentially expressed genes were identified using a threshold of $|$log2 (fold change)$|$ $\ge$1 and adjusted p-value 0.05.

RNA sequencing data for both LUAD and LSCC tissues were retrieved from the TCGA database [41]. Co-correlation patterns between ALKBH4 and other genes were evaluated using Pearson’s correlation analysis [42]. Gene Set Enrichment Analysis (GSEA) was performed with the GSEA portal (https://www.gsea-msigdb.org/gsea/index.jsp) with the “C5.go.bp.v7.5.symbols.gmt” gene set collection. The selection criteria were as follows: nominal p-value 0.05, false discovery rate (FDR) 0.25, and gene set size $>$100. Target gene prediction for 5^′-tsRNA^Glu was conducted using the TargetRank database (https://www.targetscan.org/vert_80/) [43]. The interaction probabilities between 5^′-tsRNA^Glu and candidate proteins were forecast using RPIseq (http://pridb.gdcb.iastate.edu/RPISeq/). Molecular docking simulations of 5^′-tsRNA^Glu and RPL17 were conducted using the HDOCK server (http://hdock.phys.hust.edu.cn/).

Statistical analyses were performed using SPSS software (version 22.0, SPSS Inc., IL, USA). Comparisons between two groups were assessed using unpaired two-tailed Student’s t-test, while comparisons among multiple groups were performed using one-way ANOVA. p-value 0.05 was considered statistically significant.

5^′-tsRNA^Glu (36-nt in length), is produced from the specific cleavage of the anticodon loop of tRNA^Glu(CTC). The type of 5^′-tsRNA is also addressed by another name, i.e., tiRNA, in response to stress conditions. To test if 5^′-tsRNA^Glu was induced by oxidative stress, we compared the levels of 5^′-tsRNA^Glu in NSCLC cells responding to H₂O₂ treatment. H1299 cells subjected to H₂O₂ treatment exhibited a significant increase in the levels of 5^′-tsRNA^Glu, ALKBH4, and ANG (Fig. 1A,C). Similarly, their expression levels were remarkably upregulated in A549 cells following the H₂O₂ treatment (Fig. 1B,D). Oxidative stress increased the levels of 5^′-tsRNA^Glu in NSCLC cells. Under the same conditions, the expression of ALKBH4 and ANG was also elevated, suggesting that these factors are responsive to oxidative stress.

Fig. 1.

The levels of 5^′-tsRNA^𝐆𝐥𝐮, ALKBH4 and ANG in NSCLC cells treated with H₂O₂. (A,B) The levels of 5^′-tsRNA^Glu, ALKBH4 and ANG in H1299 cells (A) and A549 cells (B) treated with H₂O₂ were measured using quantitative RT-PCR, with U6 and GAPDH as the loading controls, respectively. (C,D) The 5^′-tsRNA^Glu levels in H1299 cells (C) and A549 cells (D) treated with H₂O₂ were measured using Northern blot. The overall band intensity in PAGE served as the loading control. Data are shown as the mean $\pm$ SD (one-way ANOVA, ** p 0.01, *** p 0.001 and **** p 0.0001). Note: tsRNA-Glu is the abbreviation of 5^′-tsRNA^Glu.

RNA sequencing data for LUAD and LSCC were downloaded from the TCGA database and then used for gene expression correlation analysis, further creating a set of genes that were highly correlated with ALKBH4 expression. GSEA revealed that the gene set correlated with ALKBH4 was remarkably enriched in ‘tRNA processing’ and ‘RNA methylation’ (Supplementary Fig. 1), thus indicating that ALKBH4 might mediate the dynamic methylation of tRNAs.

Furthermore, NSCLC cells were transfected with the ALKBH4 plasmid (Fig. 2A,C). Both H1299 cells and A549 cells with ectopic expression of ALKBH4 showed a significant increase in 5^′-tsRNA^Glu levels (Fig. 2B,D). These results suggest that ALKBH4 may induce the biogenesis of 5^′-tsRNA^Glu in NSCLC cells, potentially through its tRNA demethylase activity. However, to determine whether ALKBH4 directly induces the biogenesis of 5^′-tsRNA^Glu, further studies will need to evaluate 5^′-tsRNA^Glu levels in ALKBH4 knockout cells.

Fig. 2.

The 5^′-tsRNA^𝐆𝐥𝐮 levels in NSCLC cells overexpressing ALKBH4. The ALKBH4 levels in H1299 cells (A) and A549 cells (C) transiently transfected with the ALKBH4 plasmid were determined using Western blot. GAPDH served as the loading control. The 5^′-tsRNA^Glu levels in H1299 cells (B) and A549 cells (D) transiently transfected with the ALKBH4 plasmid were determined using Northern blot. The overall band intensity in PAGE served as the loading control. The right panel shows the densitometric quantification of 5^′-tsRNA^Glu levels, presented as fold change relative to the control group. Data are shown as the mean $\pm$ SD from three independent experiments (two-sided independent Student’s t-test, * p 0.05). Note: tsRNA-Glu is the abbreviation of 5^′-tsRNA^Glu.

Next, we ivestigated the effects of ALKBH4 and 5^′-tsRNA^Glu on the proliferative ability of NSCLC cells. H1299 cells exhibiting ectopic ALKBH4 expression showed a remarkable reduction in cell proliferation, colony formation, and cell cycle arrest at the S-phase (Fig. 3). Likewise, H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics revealed decreased capabilities for cell proliferation, colony formation, and cell cycle arrest at the G2/M phase (Fig. 4A–E). Conversely, the knockdown of 5^′-tsRNA^Glu using its specific ASO significantly promoted the cell proliferation and colony formation of H1299 cells (Supplementary Fig. 2). Furthermore, the reduction in cell proliferation due to ALKBH4 overexpression was rescued by the knockdown of 5^′-tsRNA^Glu (Fig. 4F). Collectively, these results suggest that ALKBH4 inhibited NSCLC cell proliferation probably via the induction of 5^′-tsRNA^Glu.

Fig. 3.

The effects of ALKBH4 on cell proliferation and cell cycle of H1299 cells. (A) Proliferation curves of H1299 cells transiently transfected with the ALKBH4 plasmid or the pCDH vector were monitored every 24 h, over 5 days. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, ** p 0.01 and *** p 0.001). (B) Representative images of the colony formation of H1299 cells transiently transfected with the ALKBH4 plasmid or the pCDH vector were photographed (left). Their corresponding colony numbers were enumerated (right). Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, ** p 0.01). (C) The cell cycle of H1299 cells transiently transfected with the ALKBH4 plasmid or the pCDH vector was detected using flow cytometry. (D) The CyclinD1 level in H1299 cells transiently transfected with the ALKBH4 plasmid was determined using Western blot. $\beta$-actin served as the loading control.

Fig. 4.

The effects of 5^′-tsRNA^𝐆𝐥𝐮 on cell proliferation and cell cycle of H1299 cells. (A) The 5^′-tsRNA^Glu level in H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics was determined using Northern blot. The overall band intensity in PAGE served as the loading control. (B) Proliferation curves of H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics or NC mimics were monitored every 24 h, over 5 days. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, *** p 0.001). (C) Representative images of the colony formation of H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics or NC mimics were photographed (top). Their corresponding colony numbers were enumerated (bottom). Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, * p 0.05). (D) The cell cycle of H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics or NC mimics was detected using flow cytometry. (E) The levels of CyclinB1, phospho-H3 and H3 in H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics were determined using Western blot. GAPDH served as the loading control. (F) H1299 cells with ectopic expression of ALKBH4 were transiently transfected with the 5^′-tsRNA^Glu ASO or NC ASO, and proliferation curves were monitored every 24 h, over 6 days. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, * p 0.05, ** p 0.01, and *** p 0.001). Note: tsRNA-Glu is the abbreviation of 5^′-tsRNA^Glu.

We further evaluated the effects of ALKBH4 and 5^′-tsRNA^Glu on protein synthesis in NSCLC cells. OPP protein synthesis assays showed that the overexpression of ALKBH4 decreased nascent protein synthesis in H1299 cells (Fig. 5A). Furthermore, the overexpression of ALKBH4 reduced 80S monosome and polysome contents in polysome profiling (Fig. 5B). Consistently, the overexpression of 5^′-tsRNA^Glu receded nascent protein synthesis as well as 80S monosome and polysome contents in H1299 cells (Fig. 5C,D). Overall, these results indicate that both ALKBH4 and 5^′-tsRNA^Glu contributed to the downregulation of global translation efficiency in NSCLC cells.

Fig. 5.

The effects of ALKBH4 and 5^′-tsRNA^𝐆𝐥𝐮 on the global translational efficiency of H1299 cells. (A) The nascent protein synthesis of H1299 cells transiently transfected with the ALKBH4 plasmid or the pCDH vector was determined using OPP protein synthesis assay. Their relative fluorescence intensities were quantified in the right section. (B) The polysome profiling of H1299 cells transiently transfected with the ALKBH4 plasmid or the pCDH vector was analyzed. (C) The nascent protein synthesis of H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics or NC mimics was determined using OPP protein synthesis assay. Their relative fluorescence intensities were quantified in the right section. (D) The polysome profiling of H1299 cells transiently transfected with 5^′-tsRNA^Glu mimics or NC mimics was analyzed. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, ** p 0.01). Note: tsRNA-Glu is the abbreviation of 5^′-tsRNA^Glu.

Previous studies have established that tsRNAs can inhibit the translation of mRNAs via binding complementarily with the 3^′UTR of these mRNAs, in a miRNA-like manner [22, 24, 28, 44, 45, 46, 47]. Based on this, we proceeded to screen the target mRNAs of 5^′-tsRNA^Glu. The mRNA expression profiles of NSCLC cells transiently transfected with the ALKBH4 plasmid (termed the ALKBH4 profile) and those with 5^′-tsRNA^Glu mimics (termed the tsRNA-Glu profile) were detected using RNA sequencing. GSEA revealed that the biological process categories enriched in the ALKBH4 profile were translational termination, ribosome biogenesis, ncRNA processing, etc. (Supplementary Fig. 3A). The biological process categories enriched in the tsRNA-Glu profile were the regulation of response to extracellular stimulus, the negative regulation of epithelial cell proliferation, and drug catabolic process, among others (Supplementary Fig. 3B). It displayed 11 differentially expressed genes with the same trend in the ALKBH4 profile and the tsRNA-Glu profile (Supplementary Table 3). Of these, three genes that strongly associated with tumor progression, i.e., XAF1, EIF2S3B and REC8, were selected for further validation using quantitative RT-PCR. It showed that EIF2S3B and REC8 were significantly upregulated in both profiles, consistent with the RNA sequencing results, whereas XAF1 expression displayed the contrary tendency in both profiles (Supplementary Fig. 4A,B). Subsequently, we analyzed the expression patterns of XAF1 and REC8 in NSCLC tumor tissues compared to normal lung tissues, based on the TCGA and GTEx database (EIF2S3B data were unavailable). The expression levels of XAF1 and REC8 in LUAD and LSCC tissues were markedly lower than those in normal lung tissues (Supplementary Fig. 4C,D). These findings imply that REC8 mRNA may represent a downstream target of 5^′-tsRNA^Glu. Nevertheless, sequence alignment analysis failed to identify complementary pairing between 5^′-tsRNA^Glu and all regions of the REC8 mRNA, thus indicating that REC8 mRNA is probably not the direct target of 5^′-tsRNA^Glu.

Moreover, utilizing the TargetRank database, we predicted PPM1F, MECP2 and C14orf172 as the targets of 5^′-tsRNA^Glu. However, subsequent validation disproved these predictions (Supplementary Fig. 5A,B). To directly test the gene-silencing activity of 5^′-tsRNA^Glu, we next constructed a dual-luciferase reporter that has the complementary sequence of 5^′-tsRNA^Glu at the 3^′UTR of the Renilla luciferase gene (Supplementary Fig. 5C,D). The dual-luciferase reporter gene assay revealed that 5^′-tsRNA^Glu had no effect on the relative luciferase activity of the reporter gene containing the complementary sequence of 5^′-tsRNA^Glu (Supplementary Fig. 5E–H). Taking the inhibitory effects of 5^′-tsRNA^Glu on global translation efficiency into consideration, we propose that 5^′-tsRNA^Glu interacts with specific proteins rather than specific mRNAs, to suppress the global translation efficiency of NSCLC cells and thus their malignant phenotypes.

To verify the hypothesis that 5^′-tsRNA^Glu could suppress global translational efficiency via binding target proteins, we next employed RNA pull-down coupled with mass spectrometry to identify proteins that interacted with 5^′-tsRNA^Glu (Fig. 6A). Based on the results of two replicates, we identified seven overlapping proteins, including MYEF2, RPS5, RPL17, DEK, NUDT21, SRSF10, and SRSF5 (Fig. 6B). Subsequently, the interaction probabilities between 5^′-tsRNA^Glu and the candidate proteins were predicted using the RPIseq database; RPL17 exhibited the highest interaction probability among these candidate proteins (Fig. 6C). Considering that RPL17 is a ribosomal large subunit protein closely associated with translation, we selected RPL17 for further validation. Independent RNA pull-down assays and WB assays confirmed the interaction of 5^′-tsRNA^Glu and RPL17 in both H1299 and A549 cells (Fig. 6D). Consistently, RIP assays also proved that 5^′-tsRNA^Glu was enriched by the RPL17 antibody group, compared with the IgG antibody group (Fig. 6E). In addition, 5^′-tsRNA^Glu did not affect the expression of RPL17 (Supplementary Fig. 6). To gain insight into the specific binding pattern between 5^′-tsRNA^Glu and RPL17, we utilized the online database HDOCK to predict the detailed binding sites of 5^′-tsRNA^Glu to RPL17 (Fig. 6F). It showed that 5^′-tsRNA^Glu bound to the region spanning residues 137-142 of RPL17, which corresponds to the protein-rRNA interaction interface. To determine the subcellular localization of 5^′-tsRNA^Glu, we next performed nucleoplasmic separation assays and found that over 80% of 5^′-tsRNA^Glu localized in the cytoplasm of both H1299 cells and A549 cells (Fig. 6G). Furthermore, we explored the distribution of 5^′-tsRNA^Glu in the 40S, 60S, and 80S ribosomal fractions. The overexpression of 5^′-tsRNA^Glu led to an increased ratio of 5^′-tsRNA^Glu in the 60S but a decreased ratio of 5^′-tsRNA^Glu in the 80S ribosomal fraction (Fig. 6H). Given these findings, we speculate that the binding of 5^′-tsRNA^Glu to RPL17 disrupts the binding of RPL17 to rRNAs, and thus 80S ribosome assembly.

Fig. 6.

5^′-tsRNA^𝐆𝐥𝐮 selectively interacts with RPL17. (A) SDS-PAGE and then Coomassie staining of RNA pull-down samples from H1299 cells. tsGlu (RNA Probe): 5^′-tsRNA^Glu; tsGlu (DNA Probe): the corresponding DNA sequence of 5^′-tsRNA^Glu; NC: the reverse complementary DNA sequence of 5^′-tsRNA^Glu. (B) The counts of candidate 5^′-tsRNA^Glu interacted proteins identified in two biological replicates as well as these seven overlapping proteins. (C) The interaction probability of 5^′-tsRNA^Glu and candidate proteins predicted by RPIseq database. The horizontal axis represents the interaction probability using RF classifier, and the vertical axis represents the interaction probability using SVM classifier. (D) Validation of the binding of 5^′-tsRNA^Glu to RPL17 using RNA pull-down assays. 5^′-Biotinylated synthetic oligonucleotides were used for RNA pull-down assays. The antisense sequence of 5^′-tsRNA^Glu was used as the negative control. Three independent experiments were performed in triplicate. (E) Validation of the binding of 5^′-tsRNA^Glu to RPL17 using RIP assays. The 5^′-tsRNA^Glu level in the product obtained from RIP was detected using quantitative RT-PCR. Three independent experiments were performed in triplicate. (F) The binding pattern of 5^′-tsRNA^Glu and RPL17 protein predicted by HDOCK database. 5^′-tsRNA^Glu is shown in yellow, which is predicted to bind to the protein-rRNA interface (spanning residues 137–142) of RPL17. (G) The levels of 5^′-tsRNA^Glu in the cytoplasm and the nucleus of H1299 cells (left) and A549 cells (right) were determined using quantitative RT-PCR following subcellular RNA fractionation. GAPDH and U6 served as the controls for the cytoplasmic fraction and the nuclear fraction, respectively. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test. (H) Quantitative RT-PCR was utilized to detect the 5^′-tsRNA^Glu level in 40S, 60S, and 80S ribosomal fractions obtained from H1299 cells transiently transfected with NC mimics or 5^′-tsRNA^Glu mimics. Three independent experiments were performed in triplicate. Data are shown as the mean $\pm$ SD (two-sided independent Student’s t-test, * p 0.05, *** p 0.001). Note: tsRNA-Glu is the abbreviation of 5^′-tsRNA^Glu.