5′-tiRNA-Gln inhibits hepatocellular carcinoma progression by repressing translation through the interaction with eukaryotic initiation factor 4A-I

Abstract

tRNA-derived small RNAs (tsRNAs) are novel non-coding RNAs that are involved in the occurrence and progression of diverse diseases. However, their exact presence and function in hepatocellular carcinoma (HCC) remain unclear. Here, differentially expressed tsRNAs in HCC were profiled. A novel tsRNA, tRNA^Gln-TTG derived 5′-tiRNA-Gln, is significantly downregulated, and its expression level is correlated with progression in patients. In HCC cells, 5′-tiRNA-Gln overexpression impaired the proliferation, migration, and invasion in vitro and in vivo, while 5′-tiRNA-Gln knockdown yielded opposite results. 5′-tiRNA-Gln exerted its function by binding eukaryotic initiation factor 4A-I (EIF4A1), which unwinds complex RNA secondary structures during translation initiation, causing the partial inhibition of translation. The suppressed downregulated proteins include ARAF, MEK1/2 and STAT3, causing the impaired signaling pathway related to HCC progression. Furthermore, based on the construction of a mutant 5′-tiRNA-Gln, the sequence of forming intramolecular G-quadruplex structure is crucial for 5′-tiRNA-Gln to strongly bind EIF4A1 and repress translation. Clinically, 5′-tiRNA-Gln expression level is negatively correlated with ARAF, MEK1/2, and STAT3 in HCC tissues. Collectively, these findings reveal that 5′-tiRNA-Gln interacts with EIF4A1 to reduce related mRNA binding through the intramolecular G-quadruplex structure, and this process partially inhibits translation and HCC progression.

Keywords

EIF4A1 / G-quadruplex / hepatocellular carcinoma / tRNA-derived small RNA / translation initiation

1 Introduction

Primary liver cancer is the sixth most diagnosed cancer type and a major cause of cancer-related death worldwide; nearly 50% of the incidence and mortality is recorded in China [1,2]. Hepatocellular carcinoma (HCC) accounts for 75%–85% of liver cancer cases. Although surgical resection may be an effective treatment for HCC at the early stage, almost 70% of patients develop recurrent HCC after resection [3]. Moreover, considering the lack of early diagnostic marker and late symptom manifestation, the majority of patients are diagnosed with advanced HCC and have limited treatment options [4]. Therefore, the mechanism underlying progression needs to be further understood, and effective therapeutic methods need to be developed to improve the prognosis of patients with HCC.

Traditional non-coding RNAs (ncRNAs) include microRNAs (miRNAs), circular RNAs (circRNAs), and long ncRNAs (lncRNAs), which play a key role in the occurrence and development of diseases, including HCC [5–7]. Nevertheless, considering the development of high-throughput sequencing technology, tRNAs have become a source of small ncRNAs with distinct and varied functions, and they have been termed tRNA-derived small RNAs (tsRNAs) [8,9]. tsRNAs can be broadly classified into two main groups, namely, tRNA-related fragments (tRFs) generated from mature or precursor tRNA and stress-induced tRNA halves (tiRNA) generated by specific cleavage in the anticodon loops of mature tRNA, with characteristic sizes, nucleotide compositions, functions, and biogenesis.

In humans, tiRNAs are cleaved by angiogenin in the anticodon loops of mature tRNA [10,11] with length of 31–40 bases [12], leading to two subclasses of tiRNA based on whether they include the 5′ or 3′ sequence of the anticodon cleavage site. Based on the cleavage position of the mature or precursor tRNA, tRFs can be classified into at least four types, namely, 5-tRF, 3-tRF, 1-tRF, and 2-tRF [13–15], which typically have shorter sequences than tiRNAs. These tsRNAs play important roles in human cancers, suggesting that they are functional products rather than randomly degraded products [16–19]. For example, 5′-tiRNAs rather than 3′-tiRNAs promote stress granule (SG) assembly, and this process is regulated by YB-1 [20,21]. A 22-nucleotide (nt) 3′-tsRNA derived from tRNA-Leu-CAG enhances the translation of ribosomal protein mRNAs (RPS28 and RPS15), thereby promoting cell proliferation [22]. However, many tsRNAs have not been explored, thus requiring further investigation in HCC tumorigenesis.

In the present study, high-throughput sequencing was used to identify differentially expressed tsRNAs in HCC and report a novel tiRNA, namely, 5′-tiRNA-Gln, which was derived from tRNA^Gln-TTG and significantly downregulated in HCC. 5′-tiRNA-Gln was a tumor suppressor and correlated with progression in patients with HCC. Overexpressing 5′-tiRNA-Gln suppressed HCC cell growth and migration, while its knockdown resulted in their promotion. In particular, 5′-tiRNA-Gln directly bound eukaryotic initiation factor 4A-I (EIF4A1), which is the key enzyme in the translation initiation process, , causing translation repression. Furthermore, the sequence of forming G-quadruplex structure may be essential for 5′-tiRNA-Gln binding to the helicase EIF4A1 and impeding the translation of mRNAs with complex secondary structures in the 5′ untranslated regions (5′-UTR). This study provides a new target for the prognostic evaluation and treatment of HCC.

2 Materials and methods

2.1 Human samples

In total, 76 paired liver cancer specimens with clinicopathologic diagnosis were obtained from the Second Affiliated Hospital of Zhejiang University. Ethics approval was obtained from the Second Affiliated Hospital of Zhejiang University Ethics Committee. The patients’ clinical characteristics are described in Tab.1.

2.2 Small RNA high-throughput sequencing

The total RNA of the fresh tissues was qualified by agarose gel electrophoresis and quantified using a NanoDrop ND-1000 instrument (Thermo Fisher Scientific). RNA modifications that interfere with small RNA sequencing (RNA-seq) library construction were removed according to the manufacturer’s instructions of rtStar™ tRF and tiRNA pretreatment kit. Then, the sequencing library was prepared with the pretreated total RNA according to the manufacturer’s instructions of the NEBNext® Multiplex Small RNA Library Prep Set for Illumina®. For tsRNA analysis, trimmed reads were aligned, allowing for only one mismatch to the mature tRNA sequences. Reads that did not map were aligned, allowing for only one mismatch to precursor tRNA sequences by using Bowtie software. Differentially expressed tsRNAs were screened based on the count value with the R edgeR package (fold change (FC) > 1.5, P < 0.05).

2.3 Northern blotting

The total RNA was separated using 15% TBE-urea gels (Invitrogen) and visualized using SYBR Gold Nucleic Acid Gel Stain (Invitrogen) under UV illumination. Then, the separated gel was transferred to positively charged nylon membranes via semidry transfer. After UV crosslinking, the membrane was hybridized at 50 °C overnight to 5′ end-labeled antisense probes that target 5′-tiRNA-Gln. After washing, the blots were detected using a DIG Northern Starter Kit (Roche) following the manufacturer’s instructions.

2.4 RNA isolation and real-time quantitative PCR (RT-qPCR)

The total RNA of fresh tissues and cells were isolated using TRIzol (Invitrogen) and isopropanol precipitation at −20 °C overnight. RNA quality and quantity were measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Selected tiRNA expression was evaluated using TaqMan-based qPCR. RT-qPCR was performed using a Roche LightCycler® 480 II system (Roche). The thermal cycling conditions are as follows: 95 °C for 10 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 60 s. The relative tiRNA expression was normalized to human U6 levels. For mRNA transcripts, the expression level was determined using TB Green (Takara) and normalized using GAPDH. The tiRNA and mRNA primers are listed in the Supplementary Data. The FCs were calculated with the 2^−∆∆Ct method.

2.5 Cell culture and transfection

The Huh-7 cell line was cultured in DMEM supplemented with 10% fetal bovine serum. The SK-Hep-1 and Hep3B2.1-7 cell lines were cultured in MEM. The Li-7 cell line was cultured in RPMI 1640 medium. All cell lines were obtained from WuHan Procell Life Science and Technology and maintained at 37 °C with 5% CO₂. The tiRNA mimics, siRNA, and control RNA were transfected using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. A final concentration of 67 nM mimics was optimized for 35-mm culture dishes.

2.6 Cell proliferation and colony formation assay

After tiRNA or control RNA transfection, the cells were seeded in 96-well plates at an initial density of 3 × 10³ cells/well. Cell Counting Kit-8 (CCK-8) solution (10 μL) was added, and the plates were incubated at 37 °C for 2 h. Then, the absorbance was measured. For the colony formation assay, cells stably overexpressing 5′-tiRNA-Gln were plated in 6-well plates (6 × 10² cells/well) and incubated at 37 °C for 20 days. Cells were fixed with 4% paraformaldehyde for 20 min at room temperature and stained with Giemsa.

2.7 Transwell migration and invasion assays

Migration and invasion assays were performed in Transwell chambers (Corning, 8-μm pore size). A total of 8 × 10⁴ cells were resuspended in serum-free medium and seeded in the upper chamber. The cells were cultured for 24 h in 24-well plates containing 600 μL of 10% serum medium. Then, the cells were fixed in 4% paraformaldehyde and stained with crystal violet. Non-migrated cells in the upper chamber were removed using cotton swabs, while migrated cells were counted under a microscope. For the invasion assay, the upper chambers were coated with 80 μL of Matrigel plug (BD Biosciences) at a final concentration of 250 ng/mL at 37 °C for 2 h. The same steps as those for the migration assay were carried out thereafter.

2.8 Animal study

Huh-7 cells were infected with lentiviral-5′-tiRNA-Gln or lentiviral-control according to the manufacturer’s instructions (GeneChem). Huh-7 cells stably overexpressing 5′-tiRNA-Gln were selected by incubation with puromycin (3 μg/mL) for 2 days, and the results were validated by RT-qPCR. Male BALB/c nu/nu mice (4–5 weeks old, five mice per group) were injected subcutaneously with 5 × 10⁶ cells stably overexpressing 5′-tiRNA-Gln or control RNA. After 6 weeks of feeding, the mice were sacrificed. Then, the tumors were collected, measured, and weighed.

2.9 RNA pull-down and RNA immunoprecipitation (RIP)

For the RNA pull-down assay, Huh-7 cells were transfected with 3′ end biotinylated tiRNA or control RNA (Takara). After 48 h of transfection, the cells were lysed in lysis buffer for 15 min on ice. Streptavidin beads were washed thrice with RNase-free wash buffer, twice with solution A, once with solution B, and then resuspended in lysis buffer. Cell lysis buffer was added to the streptavidin beads and rotated overnight at 4 °C. The immobilized RNA–streptavidin complex was washed twice with lysis buffer. Proteins were eluted using 30 μL 1× SDS loading.

RIP assay was carried out using a Magna RIP kit (Millipore) according to the manufacturer’s instructions. Briefly, transfected cells were harvested and lysed in RIP lysis buffer on ice and stored at −80 °C. Magnetic beads were washed and resuspended in RIP wash buffer. Afterward, 5 μg rabbit anti-EIF4A1 antibody (ab31217, Abcam) or negative control immunoglobulin G (IgG) was added. Then, the RIP lysate was added to the beads–antibody complex in RIP buffer and rotated overnight at 4 °C. The bead complexes were washed and resuspended in proteinase K buffer to separate the proteins. Then, the RNA was extracted for subsequent qPCR detection.

2.10 Western blot analysis

Cells were lysed with RIPA lysis buffer. Total proteins were quantified using a bicinchoninic acid (BCA) protein assay kit (Beyotime Biotechnology). The proteins were separated by electrophoresis on 10%–15% SDS–polyacrylamide gel and transferred to nitrocellulose membranes (Whatman). The membranes were blocked with 5% bovine serum albumin (BSA) in TBST buffer, incubated with primary antibodies and then with the secondary antibody, and finally subjected to chemiluminescent detection (Bio-Rad). The antibodies used were rabbit anti-EIF4A1 (ab31217, Abcam), rabbit anti-ARAF (22129-1-AP, Proteintech), rabbit anti-BID (10988-1-AP, Proteintech), rabbit anti-MEK1 (51080-1-AP, Proteintech), mouse anti-MEK2 (67410-1-AP, Proteintech), mouse anti-p-ERK1/2 (8544, CST), mouse anti-STAT3 (60199-1-IG, Proteintech), rabbit anti-p-STAT3 (9145, CST), rabbit anti-TUBULIN (11224-1-AP, Proteintech), mouse anti-GAPDH (60004-1-IG, Proteintech), rabbit anti-ACTIN (23660-1-AP, Proteintech), and rabbit anti-puromycin (MABE341, Merck).

2.11 Puromycin capture assay and polysome profiling

Huh-7 cells at 70%–80% confluence were treated with 10 μg/mL puromycin 10 min before being harvested. After washing twice with ice-cold PBS, the cells were lysed, and proteins were detected with anti-puromycin antibody.

The transfected Huh-7 cells were treated with 100 μg/ml cycloheximide (CHX) for 5 min to freeze the translating ribosomes. Then, the cells were lysed with lysis buffer (10 mmol/L HEPES, pH 7.4, 100 mmol/L KCl, 5 mmol/L MgCl₂, 100 μg/mL CHX, 5 mmol/L DTT, 1% Triton X-100) on ice, and the supernatant was collected after centrifugation. The supernatant was layered over 10%–50% w/v sucrose density gradient and sedimented using a Beckman SW41Ti rotor at 178 305×g for 2.5 h at 4 °C. The sample was fractioned using Gradient Station (BioCamp) equipped with a UV monitor, and then collected using a fraction collector (FC203B, Gilson).

2.12 Colocalization of 5′-tiRNA-Gln and EIF4A1

Biotin-labeled 5′-tiRNA-Gln was transfected into Huh-7 cells. Hybridization was performed using an RNA FISH kit (SA-Biotin system, GenePharma) according to the manufacturer’s instructions. Then, the cells were incubated with anti-EIF4A1, and the nuclei were stained with DAPI. The staining results were observed under a fluorescence microscope (Nikon).

2.13 Mass spectrometry (MS) and bioinformatics analysis

After RNA pull-down, the protein samples were subjected to SDS-PAGE. Then, the gel was stained with Coomassie bright blue. The band was cut and analyzed by liquid chromatography-tandem MS (LC-MS/MS, Jingjie PTM Biolabs). The resulting MS/MS data were processed using MaxQuant v.1.6.6.0. Tandem mass spectra were searched against the human UniProt database concatenated with a reverse decoy database.

2.14 Immunohistochemistry (IHC) staining and score

IHC analysis was performed on a tissue microarray containing 51 paired samples of tumor and adjacent non-tumorous tissues. Briefly, the slide was blocked in 3% BSA, incubated with anti-ARAF, anti-MEK1, anti-MEK2, and anti-STAT3 primary antibodies overnight at 4 °C, incubated with horseradish peroxidase-conjugated anti-rabbit or mouse IgG secondary antibody, and visualized using diaminobenzidine tetrahydrochloride. The nuclei were labeled with hematoxylin. The IHC staining images were photographed under a microscope (Nikon). IHC scores were determined by a pathologist and identified based on the immunostaining intensity and the percentage of positively stained area (intensity score: 0, no staining; 1, weak positive; 2, moderate positive; 3, strong positive; percentage score: 1, < 25% positive cells/areas; 2, 25%–49% positive cells/areas; 3, 50%–74% positive cells/areas; 4, ≥ 75% positive cells/areas). The final score was calculated by adding the above intensity and percentage scores.

2.15 Statistical analysis

All statistical analyses and graphs were processed in GraphPad Prism 7.0. Data are presented as the mean ± standard error of the mean. The difference between two groups was assessed using Student’s t-test. Statistical significance among at least three groups was determined using one-way analysis of variance. The Spearman correlation was calculated between the 5′-tiRNA-Gln expression levels of tumors and IHC scores. Statistical significance was considered at P < 0.05, P < 0.01, and P < 0.001.

3 Results

3.1 tsRNA profile characterization of paired HCC tissues

To determine the expression of specific tsRNAs in primary HCC, we used four paired tumor and adjacent non-tumor liver tissues for the analysis of the tiRNA levels via RNA-PAGE. The HCC tissues had lower total tiRNAs than the adjacent tissues (Fig.1). Then, the small RNA transcriptome in these samples was analyzed via high-throughput sequencing. The results showed that in either tumor or adjacent non-tumor tissues, reads counts were enriched at 20–24 and 31–33 nt, referring to the position of miRNAs, tRFs, and 5′-tiRNAs (Fig.1). Based on the statistical information of the reads, miRNAs accounted for the vast majority of the total reads, followed by mature tRNA-derived small RNAs and the precursor tRNAs ( < 1%, Fig.1). Therefore, the miRNA expression profiling was also analyzed (Fig. S1A), and the exact results of all differentially expressed miRNAs are shown in the Supplementary Data.

The abundance of specific tsRNAs was evaluated using their sequencing counts and was normalized as counts per million of total aligned reads (CPM). At CPM < 20 in all samples, the tsRNAs were filtered. In total, 77 tsRNAs were found in tumor tissues, 67 tsRNAs were found in adjacent non-tumor tissues, and 208 tsRNAs were commonly expressed (Fig.1). The distribution of the proportion for each tsRNA subtype in the tumor or adjacent tissues was also analyzed. Results showed that 5-tRF accounted for the highest proportion, followed by 3-tRF and 5′-tiRNA, both in tumor and adjacent non-tumor tissues (Fig.1 and 1F). At least 500 genes that encode a set of 49 different tRNAs were identified in the human genome, indicating the existence of multiple tRNA isodecoders and isoacceptors [23,24]. The distribution of the number for each tsRNA subtype against the exact tRNA isodecoders and isoacceptors was also analyzed (Fig. S1B). Low expression level of tsRNAs was observed in HCC via RNA PAGE, and the tsRNA profile was characterized.

3.2 Selected 5′-tiRNA-Gln is downregulated in tumor tissues and correlates with HCC progression

Differentially expressed tsRNAs were analyzed to elucidate the potential implication of tsRNA expression in HCC progression. In total, 80 upregulated and 74 downregulated tsRNAs were identified (Fig.2). The Supplementary Data show the exact data of all differentially expressed tsRNAs. The function of tiRNAs in HCC was the focus of our attention, and 19 tiRNAs from the 154 differentially expressed tsRNAs were found (Fig.2). These 19 tiRNAs were searched both in the tRFdb [25] and MINTbase [26], and only nine tiRNAs were annotated in the MINTbase (Fig. S2A). Then, the expression data of the four highest changed tiRNAs were obtained from the MINTbase. Only the expression of tRF-34-V29K9UV36562FQ (5′-tiRNA-Gln) from the database was in accordance with our data (P = 0.09, Fig. S2B). 5′-tiRNA-Gln was derived from the tRNA^Gln-TTG, containing the D-loop and partial anticodon loop (Fig. S2C). The expression level of 5′-tiRNA-Gln by RT-qPCR in an expanded 76 HCC tissue pairs was verified. Results show that consistent with the differential analyses results, 5′-tiRNA-Gln was significantly downregulated in tumor tissues, suggesting a tumor suppressor role (P < 0.001, Fig.2). The Northern blot experiment yielded similar results (Fig.2).

Correlation analyses were carried out to determine the role of 5′-tiRNA-Gln expression in patients with HCC. Tumor size and distal metastasis were correlated with 5′-tiRNA-Gln, whereas gender, age, hepatitis B infection, alpha-fetoprotein level, liver cirrhosis, tumor number, and differentiation were not (Table S1). The metastasis group had a significantly lower relative mean level of 5′-tiRNA-Gln than the non-metastasis group (P = 0.04, Fig.2). The receiver-operating characteristic (ROC) analysis showed that 5′-tiRNA-Gln could distinguish HCC metastasis group from non-metastasis group with an area under the curve (AUC) of 0.673 (95% CI: 0.551–0.794, P < 0.01; Fig. S3A). According to the tumor size, the patients were divided into tumor ≥ 5 cm and < 5 cm groups, and decreased 5′-tiRNA-Gln expression was associated positively with large tumors (P = 0.04, Fig.2). Spearman’s correlation analysis showed a significantly negative correlation between the tumor size and the expression of 5′-tiRNA-Gln (r = −0.2325, P < 0.05; Fig. S3B). These results indicate that 5′-tiRNA-Gln was significantly downregulated in HCC and plays a negative regulatory role in HCC metastasis and tumor size.

3.3 overexpression inhibits HCC cell proliferation, migration, and invasion

Considering that 5′-tiRNA-Gln was enriched in the adjacent nontumor tissues and correlated with HCC progression, altering its levels might be crucial for cell activity. Accordingly, its expression levels in Huh-7, SK-Hep-1, Hep3B2.1-7, and Li-7 cells were examined via RT-qPCR (Fig. S4A). Then, it was overexpressed in Huh-7 cell lines and verified through 15% urea gel electrophoresis (Fig. S4B). After 24 h of transfection of 5′-tiRNA-Gln mimic, its effects on HCC cell growth, migration, and invasion were analyzed. The CCK-8 assay revealed that the mimic decreased cell viability (Fig.3). Transwell analyses showed that the overexpression of the mimic decreased the Huh-7 cell migration and invasive capacities (Fig.3).

Loss-of-function experiments were performed via the siRNA approach. SK-Hep-1 cells were transfected with siRNA targeting 5′-tiRNA-Gln by Lipofectamine 3000 for 24 h. Then, the cell viability, migration, and invasion were investigated (Fig.3 and 3D). The knockdown of 5′-tiRNA-Gln enhanced SK-Hep-1 cell viability, migration, and invasion. To determine whether the effect was induced by the loss of 5′-tiRNA-Gln rather than mature tRNA-Gln, we detected their expression levels via Northern blot analysis and RT-qPCR (Fig. S4C). The siRNA specifically inhibited most of the tiRNA expression but did not affect its mature tRNA levels. For investigating the 5′-tiRNA-Gln function in an in vivo tumorigenesis assay, we constructed a stable overexpression 5′-tiRNA-Gln Huh-7 cell line and tested the expression levels (Fig. S4D). In the subcutaneous tumor xenograft model, Huh-7 cells stably overexpressing 5′-tiRNA-Gln or control tiRNA were transplanted into nude mice for 6 weeks. The 5′-tiRNA-Gln tumors had reduced weight compared with the control (Fig.3). The colony formation assay yielded similar results (Fig.3). These results support our hypothesis that 5′-tiRNA-Gln inhibits HCC cell activity.

3.4 selectively interacts with EIF4A1

To explore the potential molecular mechanisms of 5′-tiRNA-Gln in HCC, we transfected biotinylated 5′-tiRNA-Gln or control RNA into Huh-7 cells for 48 h. Then, the RNA–protein complexes were immobilized on streptavidin beads. MS identified 84 differential proteins with specific binding to 5′-tiRNA-Gln rather than the control RNA. Based on the protein score, a eukaryotic translation initiation factor (EIF4A1) was selected as the candidate (Fig.4), and it is a key enzyme in the translation initiation process. Based on the survival curve analysis from GEPIA, higher EIF4A1 expression levels in HCC imply lower survival (Fig.4), indicating an oncoprotein role. Independent RNA pull-down experiments confirmed that EIF4A1 bound to 5′-tiRNA-Gln more strongly than to the control RNA (Fig.4). RIP assay confirmed that 5′-tiRNA-Gln was significantly enriched in EIF4A1 immunoprecipitate compared with the IgG control (Fig.4). RNA FISH and immunofluorescence assays revealed that exogenous 5′-tiRNA-Gln was located in both the nucleus and cytoplasm of Huh-7 cells and confirmed the occurrence of 5′-tiRNA-Gln and EIF4A1 colocalization in the cytoplasm (Fig.4). Therefore, 5′-tiRNA-Gln suppresses HCC tumors presumably by interacting with EIF4A1 in the cytoplasm.

3.5 5′-tiRNA-Gln overexpression represses translation

Based on the biological function of EIF4A1 in eukaryotic translation initiation, the effect of 5′-tiRNA-Gln overexpression on EIF4A1-mediated mRNA translation was investigated. First, EIF4A1 expression levels were detected, and the results show that EIF4A1 quantity did not change with the transfection of the control RNA or 5′-tiRNA-Gln (Fig. S5A). Then, the effect of 5′-tiRNA-Gln on EIF4A1 function through binding via puromycin capture was determined to identify the nascent protein synthesis level (Fig.5). Results show that 5′-tiRNA-Gln inhibited the protein synthesis rate. Specific differentially expressed proteins were determined via proteomics assay, from which 6,490 proteins were identified, including 284 significantly downregulated and 55 upregulated ones compared with the control group (Fig.5, Fig. S5B). Therefore, 5′-tiRNA-Gln overexpression partially repressed Huh-7 cell translation.

The Clusters of Orthologous Groups/Eukaryotic Orthologous Groups (COG/KOG) functional classification analysis showed that the downregulated proteins were mainly associated with the processes of transcription and translation, ribosomal structure, and biogenesis (Fig.5), which was vital for cell survival. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of the downregulated proteins showed that some pathways were changed significantly, including that for hepatitis B, MAPK signaling, hepatitis C, and pathways in cancer (Fig.5).

Persistent infections with HBV or HCV accounted for the majority of cases of HCC in China. Therefore, five candidate proteins (ARAF, MEK1, MEK2, BID, and STAT3) were used for further verification from the intersection of hepatitis B, hepatitis C, and pathway in cancer in the KEGG enrichment analysis (Fig. S5C). Based on the results of Western blot analysis, the expression of ARAF, MEK1/2, and STAT3 was downregulated by overexpressing 5′-tiRNA-Gln and the downstream phosphorylated proteins (p-ERK1/2 and p-STAT3). However, no noticeable difference was observed in the expression of BID (Fig.5).

These results indicate that 5′-tiRNA-Gln overexpression directly targets EIF4A1 function rather than its expression level, thereby repressing EIF4A1 related translation.

3.6 The G-quadruplex structure is necessary for 5′-tiRNA-Gln

EIF4A1 or ATP-dependent RNA helicase EIF4A1, is a key protein of the eukaryotic translation initiation process that unwinds complex RNA secondary structures in the 5′-UTR of mRNAs, and this process is required for ribosome binding and subsequent scanning for the initiator codon [27,28]. The literature review revealed that EIF4A1-mediated helicase activity is associated with the translation of GC-rich mRNAs, which can form multiple G-quadruplexes [29–32]. Therefore, 5′-tiRNA-Gln may serve as a G-quadruplex to regulate EIF4A1 activity. Accordingly, QGRS Mapper, a web-based server [33], was used to predict the potential ability of 5′-tiRNA-Gln to form G-quadruplexes. Results show that one G-quadruplex was formed from 5′-tiRNA-Gln and contained 19 overlaps (Fig. S6A). To confirm our hypothesis, we constructed mutant 5′-tiRNA-Gln without the ability to form a G-quadruplex based on QGRS Mapper (Fig. S6B). Huh-7 cell viability, migration, and invasive ability were investigated by overexpressing the mutant 5′-tiRNA-Gln (Fig.6 and 6B). No significant difference was observed between the mutant and control RNA, but it was significantly different from 5′-tiRNA-Gln.

Then, biotin-labeled control RNA, mutant 5′-tiRNA-Gln, and 5′-tiRNA-Gln were transfected into Huh-7 cells for 48 h to pull down EIF4A1. The control RNA and mutant 5′-tiRNA-Gln failed to pull down EIF4A1 specifically (Fig.6). Next, the ability of mutant 5′-tiRNA-Gln to repress EIF4A1-mediated translation was determined. As an eIF4F complex subunit, EIF4A1 participates in translation initiation by unwinding the mRNA secondary structure and recruiting ribosomes to start translation [27,28]. Therefore, the polysome profile was determined, and the results show that 5′-tiRNA-Gln reduced mRNA binding to polyribosomes compared with the control or mutant (Fig.6).

Further, we validated whether the expression of ARAF, MEK1/2, and STAT3 would be downregulated by the mutant 5′-tiRNA-Gln. Their mRNA levels and protein quantity were detected. The ARAF, MEK1/2, and STAT3 cytoplasmic mRNAs were not significantly different (Fig.6), while the overexpressed 5′-tiRNA-Gln, rather than the control or mutant, significantly reduced their mRNA binding to EIF4A1 (Fig.6) and their protein levels (Fig.6). These results suggest that the potential of 5′-tiRNA-Gln to form G-quadruplexes is necessary for its regulation of EIF4A1-mediated translation and inhibiting HCC progression.

3.7 5′-tiRNA-Gln expression level in HCC tumors is negatively correlated with ARAF, MEK1/2 or STAT3 immuno-expression

The 5′-tiRNA-Gln expression downregulated ARAF, MEK1/2, and STAT3 protein levels in Huh-7 cells. Accordingly, the association between 5′-tiRNA-Gln and ARAF, MEK1/2, or STAT3 expression levels in clinical HCC tissues was determined. ARAF, MEK1/2, and STAT3 are oncogenes that participate in tumor-promoting signaling pathway [34–36] and are correlated with the prognosis of patients with HCC. In the present study, IHC assays showed that HCC tissues had significantly elevated ARAF, MEK1/2, and STAT3 immuno-expression levels compared with the adjacent tissues (Fig. S7A and S7B). Furthermore, Spearman’s correlation analysis of the expression level between 5′-tiRNA-Gln and ARAF, MEK1/2, or STAT3 in the HCC tissues revealed a significant negative correlation (Fig.7). Therefore, 5′-tiRNA-Gln acts as a tumor suppressor and is negatively correlated with ARAF, MEK1/2, and STAT3 expression in HCC.

4 Discussion

tRNAs, which are typically 76–90-nt long, are well known for their primary ability to transport amino acids during protein translation. With the wide application of next-generation sequencing, tRNAs have also been identified as the source of novel small ncRNAs (i.e., tsRNAs), which are dysregulated in multiple human diseases, including tumors [16,17,37]. Although tsRNAs play a molecular function in carcinogenesis, considering the large diversity of tRNA genes in the human genome, a large variety of tsRNAs remains to be discovered and studied. In the present study, a tsRNA library of patients with HCC was constructed, and differentially expressed tsRNAs were analyzed. Based on the results, 5′-tiRNA-Gln was downregulated in almost all tumor tissues compared with that in paired normal liver tissues and was significantly associated with tumor size and metastasis. 5′-tiRNA-Gln gain- or loss-of-function experiments in HCC cells demonstrated that 5′-tiRNA-Gln regulates HCC cell proliferation, migration, and invasion in accordance with the clinical analysis results. Further, the subcutaneous tumor xenograft model showed that 5′-tiRNA-Gln overexpression suppressed tumor formation in vivo. Therefore, 5′-tiRNA-Gln may act as a tumor suppressor in HCC.

tRNA derivate fragments are engaged in the response to cellular stress, especially 5′-tiRNAs, by targeting RNA binding proteins (RBPs) [38,39]. Shorter tRFs may act similarly to canonical miRNAs by directly targeting mRNA [40–43]. One alternative mechanism of tRFs is the promotion of the translation of certain RPS mRNAs by binding with the stem-loop structure [22]. Therefore, tiRNAs may act by binding to proteins, while tRFs bind to mRNAs. Accordingly, the RBPs of 5′-tiRNA-Gln were identified through the protein spectrum. EIF4A1 was selected and was negatively associated with the survival of patients with HCC. RNA pull-down, RIP, and fluorescence colocalization assays confirmed that 5′-tiRNA-Gln bound with EIF4A1 in the cytoplasm. Therefore, 5′-tiRNA-Gln may inhibit HCC cell proliferation, migration, and invasion probably by directly interacting with EIF4A1.

EIF4A1 plays an important role in initiating translation in eukaryotic cells by assembling EIF4F in combination with EIF4G and EIF4E [27,28]. Therefore, 5′-tiRNA-Gln might repress EIF4A1-mediated translation. Puromycin assay results confirmed that 5′-tiRNA-Gln overexpression partially inhibited the translation of Huh-7 cells. The polysome profile showed that 5′-tiRNA-Gln overexpression reduced mRNA binding to polyribosomes. Proteomics assay was carried out to determine the exact proteins affected by 5′-tiRNA-Gln. The majority of the differentially expressed proteins were downregulated (5 FCs more than the upregulated proteins). KEGG pathway enrichment analyses of the downregulated proteins showed that hepatitis B, hepatitis C, and pathway in cancer were affected. Based on their intersection, ARAF, MEK1/2, STAT3, and BID were selected as candidates for further verification by Western blot analysis. Except for BID, the expression level of ARAF, MEK1/2, and STAT3 was reduced by overexpressing 5′-tiRNA-Gln. RIP assays revealed that the mRNA binding to EIF4A1 was significantly downregulated by overexpressing 5′-tiRNA-Gln, resulting in reduced protein levels after translation. The downstream phosphorylated proteins (p-ERK1/2 and p-STAT3) were also reduced. These results suggest that 5′-tiRNA-Gln suppresses hepatoma cells presumably by repressing the EIF4A1-mediated translation of the related proteins.

The specific molecular mechanisms in which 5′-tiRNA-Gln interacts with EIF4A1 and subsequently affects translation remain unknown. Notably, 5′-tiRNAs that consists of the 5′-terminal oligoguanine (5′-TOG) motif (e.g., 5′-tiRNA-Ala, 5′-tiRNA-Cys) can fold into intermolecular G-quadruplex-like structures and interact with eIF4G, thereby displacing eIF4F from the m⁷GTP cap of mRNAs and triggering phosphorylated eIF2α-independent SG assembly [21,44–46]. However, 5′-tiRNA-Gln does not contain the 5′-TOG motif and binds with EIF4A1, thus requiring further investigation. EIF4A1 participates in translating mRNAs with long and complex 5′-UTRs, including computationally predicted G-quadruplex structures, because of its RNA helicase activity [29,47,48]. RNA G-quadruplexes are guanine-rich nucleic acid sequences that can form four-stranded structures [49]. A minimum number of 12 nt (e.g., (CGG)₄) is required to form two stacks of four guanosines with non-Watson–Crick interactions (Hoogsteen hydrogen bonds) [29,50]. Accordingly, we examined whether 5′-tiRNA-Gln could serve as an intramolecular G-quadruplex to partially repress EIF4A1-dependent translation by competitively binding EIF4A1. By using QGRS Mapper, a G-quadruplex prediction website, we found that 5′-tiRNA-Gln can form a G-quadruplex containing 19 overlaps. To verify our hypothesis, we constructed a mutant 5′-tiRNA-Gln that failed to form a G-quadruplex and repress the EIF4A1-mediated translation compared with 5′-tiRNA-Gln. Importantly, RNA pull-down and RIP assays demonstrated that the mutant 5′-tiRNA-Gln failed to pull down EIF4A1; and was not different in the related mRNA binding to EIF4A1 with the control RNA, but was different from that of 5′-tiRNA-Gln. Moreover, ARAF, MEK1/2, and STAT3 5′-UTRs were rich in G-quadruplex structures (Fig. S6C–S6F). IHC assay revealed a negative relationship between the expression of 5′-tiRNA-Gln and ARAF, MEK1/2, or STAT3. These results suggest that 5′-tiRNA-Gln acts as a tumor suppressor, and its predicted intramolecular G-quadruplex structure is essential for repressing translation probably by competitively binding more EIF4A1 than the mRNAs whose 5′-UTRs contain rich G-quadruplex structures (Fig.7).

However, the present study has some limitations. Instead of being verified experimentally, the G-quadruplex structure was predicted in silico. Alternatively, we proved that the sequence was needed to form a G-quadruplex structure of 5′-tiRNA-Gln. Several studies have reported the possibility of the use of tsRNAs as potential diagnostics markers in multiple human cancers, including HCC [51–54]. However, 5′-tiRNA-Gln expression levels were only detected in HCC tissues and not in the circulation. Additionally, the normal populations have not been obtained in our research, preventing us from evaluating the diagnostic value of 5′-tiRNA-Gln. These topics will be the subjects of our further research. Due to their small size, tsRNAs have great potential as therapeutic targets as small molecules [55]. For example, Hani Goodarzi et al. reported that the transfection of mimetics of YBX1 binding tsRNAs can significantly inhibit lung metastasis in vivo [14], indicating their potential of suppressing tumor progression and metastasis. In the present study, 5′-tiRNA-Gln acts as a tumor suppressor in HCC, thus providing the possibility to verify the potential of 5′-tiRNA-Gln as a therapeutic target through animal experiments and clinical trials in the future.

In summary, a panel of differentially expressed tsRNAs in HCC was identified, and a novel tiRNA, namely, 5′-tiRNA-Gln, which serves as a tumor suppressor. Was characterized. 5′-tiRNA-Gln inhibited HCC cell proliferation and mobility presumably by negatively regulating the key enzyme in the translation initiation process, EIF4A1, which caused the alteration of proteins in multiple signaling pathways. Moreover, the sequence of forming an intramolecular G-quadruplex structure is the basis for strong binding of EIF4A1.

References

[1] Zhou M, Wang H, Zeng X, Yin P, Zhu J, Chen W, Li X, Wang L, Wang L, Liu Y, Liu J, Zhang M, Qi J, Yu S, Afshin A, Gakidou E, Glenn S, Krish VS, Miller-Petrie MK, Mountjoy-Venning WC, Mullany EC, Redford SB, Liu H, Naghavi M, Hay SI, Wang L, Murray CJL, Liang X. Mortality, morbidity, and risk factors in China and its provinces, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2019; 394(10204): 1145–1158

[2] Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021; 71(3): 209–249

[3] Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nat Rev Gastroenterol Hepatol 2019; 16(10): 589–604

[4] Kobayashi T, Aikata H, Kobayashi T, Ohdan H, Arihiro K, Chayama K. Patients with early recurrence of hepatocellular carcinoma have poor prognosis. Hepatobiliary Pancreat Dis Int 2017; 16(3): 279–288

[5] Wang J, Zhu S, Meng N, He Y, Lu R, Yan GR. ncRNA-encoded peptides or proteins and cancer. Mol Ther 2019; 27(10): 1718–1725

[6] Wong CM, Tsang FHC, Ng IOL. Non-coding RNAs in hepatocellular carcinoma: molecular functions and pathological implications. Nat Rev Gastroenterol Hepatol 2018; 15(3): 137–151

[7] Klingenberg M, Matsuda A, Diederichs S, Patel T. Non-coding RNA in hepatocellular carcinoma: mechanisms, biomarkers and therapeutic targets. J Hepatol 2017; 67(3): 603–618

[8] Su Z, Wilson B, Kumar P, Dutta A. Noncanonical roles of tRNAs: tRNA fragments and beyond. Annu Rev Genet 2020; 54(1): 47–69

[9] Kim HK, Yeom JH, Kay MA. Transfer RNA-derived small RNAs: another layer of gene regulation and novel targets for disease therapeutics. Mol Ther 2020; 28(11): 2340–2357

[10] Ivanov P, Emara MM, Villen J, Gygi SP, Anderson P. Angiogenin-induced tRNA fragments inhibit translation initiation. Mol Cell 2011; 43(4): 613–623

[11] Yamasaki S, Ivanov P, Hu GF, Anderson P. Angiogenin cleaves tRNA and promotes stress-induced translational repression. J Cell Biol 2009; 185(1): 35–42

[12] Li S, Hu GF. Emerging role of angiogenin in stress response and cell survival under adverse conditions. J Cell Physiol 2012; 227(7): 2822–2826

[13] Lee YS, Shibata Y, Malhotra A, Dutta A. A novel class of small RNAs: tRNA-derived RNA fragments (tRFs). Genes Dev 2009; 23(22): 2639–2649

[14] Goodarzi H, Liu X, Nguyen HC, Zhang S, Fish L, Tavazoie SF. Endogenous tRNA-derived fragments suppress breast cancer progression via YBX1 displacement. Cell 2015; 161(4): 790–802

[15] Zheng LL, Xu WL, Liu S, Sun WJ, Li JH, Wu J, Yang JH, Qu LH. tRF2Cancer: a web server to detect tRNA-derived small RNA fragments (tRFs) and their expression in multiple cancers. Nucleic Acids Res 2016; 44(W1): W185–93

[16] Zeng T, Hua Y, Sun C, Zhang Y, Yang F, Yang M, Yang Y, Li J, Huang X, Wu H, Fu Z, Li W, Yin Y. Relationship between tRNA-derived fragments and human cancers. Int J Cancer 2020; 147(11): 3007–3018

[17] Zhu L, Ge J, Li T, Shen Y, Guo J. tRNA-derived fragments and tRNA halves: the new players in cancers. Cancer Lett 2019; 452: 31–37

[18] Magee R, Rigoutsos I. On the expanding roles of tRNA fragments in modulating cell behavior. Nucleic Acids Res 2020; 48(17): 9433–9448

[19] Chen Q, Zhang X, Shi J, Yan M, Zhou T. Origins and evolving functionalities of tRNA-derived small RNAs. Trends Biochem Sci 2021; 46(10): 790–804

[20] Emara MM, Ivanov P, Hickman T, Dawra N, Tisdale S, Kedersha N, Hu GF, Anderson P. Angiogenin-induced tRNA-derived stress-induced RNAs promote stress-induced stress granule assembly. J Biol Chem 2010; 285(14): 10959–10968

[21] Lyons SM, Achorn C, Kedersha NL, Anderson PJ, Ivanov P. YB-1 regulates tiRNA-induced stress granule formation but not translational repression. Nucleic Acids Res 2016; 44(14): 6949–6960

[22] Kim HK, Fuchs G, Wang S, Wei W, Zhang Y, Park H, Roy-Chaudhuri B, Li P, Xu J, Chu K, Zhang F, Chua MS, So S, Zhang QC, Sarnow P, Kay MA. A transfer-RNA-derived small RNA regulates ribosome biogenesis. Nature 2017; 552(7683): 57–62

[23] Geslain R, Pan T. Functional analysis of human tRNA isodecoders. J Mol Biol 2010; 396(3): 821–831

[24] Goodenbour JM, Pan T. Diversity of tRNA genes in eukaryotes. Nucleic Acids Res 2006; 34(21): 6137–6146

[25] Kumar P, Mudunuri SB, Anaya J, Dutta A. tRFdb: a database for transfer RNA fragments. Nucleic Acids Res 2015; 43(D1): D141–D145

[26] Pliatsika V, Loher P, Magee R, Telonis AG, Londin E, Shigematsu M, Kirino Y, Rigoutsos I. MINTbase v2.0: a comprehensive database for tRNA-derived fragments that includes nuclear and mitochondrial fragments from all The Cancer Genome Atlas projects. Nucleic Acids Res 2018; 46(D1): D152–D159

[27] Hinnebusch AG. The scanning mechanism of eukaryotic translation initiation. Annu Rev Biochem 2014; 83(1): 779–812

[28] Jackson RJ, Hellen CUT, Pestova TV. The mechanism of eukaryotic translation initiation and principles of its regulation. Nat Rev Mol Cell Biol 2010; 11(2): 113–127

[29] Wolfe AL, Singh K, Zhong Y, Drewe P, Rajasekhar VK, Sanghvi VR, Mavrakis KJ, Jiang M, Roderick JE, Van der Meulen J, Schatz JH, Rodrigo CM, Zhao C, Rondou P, de Stanchina E, Teruya-Feldstein J, Kelliher MA, Speleman F, Porco JA Jr, Pelletier J, Rätsch G, Wendel HG. RNA G-quadruplexes cause eIF4A-dependent oncogene translation in cancer. Nature 2014; 513(7516): 65–70

[30] Modelska A, Turro E, Russell R, Beaton J, Sbarrato T, Spriggs K, Miller J, Gräf S, Provenzano E, Blows F, Pharoah P, Caldas C, Le Quesne J. The malignant phenotype in breast cancer is driven by eIF4A1-mediated changes in the translational landscape. Cell Death Dis 2015; 6(1): e1603

[31] Singh K, Lin J, Lecomte N, Mohan P, Gokce A, Sanghvi VR, Jiang M, Grbovic-Huezo O, Burčul A, Stark SG, Romesser PB, Chang Q, Melchor JP, Beyer RK, Duggan M, Fukase Y, Yang G, Ouerfelli O, Viale A, de Stanchina E, Stamford AW, Meinke PT, Rätsch G, Leach SD, Ouyang Z, Wendel HG. Targeting eIF4A-dependent translation of KRAS signaling molecules. Cancer Res 2021; 81(8): 2002–2014

[32] Iwasaki S, Iwasaki W, Takahashi M, Sakamoto A, Watanabe C, Shichino Y, Floor SN, Fujiwara K, Mito M, Dodo K, Sodeoka M, Imataka H, Honma T, Fukuzawa K, Ito T, Ingolia NT. The translation inhibitor rocaglamide targets a bimolecular cavity between eIF4A and polypurine RNA. Mol Cell 2019; 73(4): 738–748.e9

[33] Kikin O, D’Antonio L, Bagga PS. QGRS Mapper: a web-based server for predicting G-quadruplexes in nucleotide sequences. Nucleic Acids Res 2006; 34(suppl_2): W676–W682

[34] Calvisi DF, Ladu S, Gorden A, Farina M, Conner EA, Lee JS, Factor VM, Thorgeirsson SS. Ubiquitous activation of Ras and Jak/Stat pathways in human HCC. Gastroenterology 2006; 130(4): 1117–1128

[35] Delire B, Stärkel P. The Ras/MAPK pathway and hepatocarcinoma: pathogenesis and therapeutic implications. Eur J Clin Invest 2015; 45(6): 609–623

[36] Ranjpour M, Wajid S, Jain SK. Elevated expression of A-Raf and FA2H in hepatocellular carcinoma is associated with lipid metabolism dysregulation and cancer progression. Anticancer Agents Med Chem 2019; 19(2): 236–247

[37] Yu M, Lu B, Zhang J, Ding J, Liu P, Lu Y. tRNA-derived RNA fragments in cancer: current status and future perspectives. J Hematol Oncol 2020; 13(1): 121

[38] Han L, Lai H, Yang Y, Hu J, Li Z, Ma B, Xu W, Liu W, Wei W, Li D, Wang Y, Zhai Q, Ji Q, Liao T. A 5′-tRNA halve, tiRNA-Gly promotes cell proliferation and migration via binding to RBM17 and inducing alternative splicing in papillary thyroid cancer. J Exp Clin Cancer Res 2021; 40(1): 222

[39] Saikia M, Jobava R, Parisien M, Putnam A, Krokowski D, Gao XH, Guan BJ, Yuan Y, Jankowsky E, Feng Z, Hu GF, Pusztai-Carey M, Gorla M, Sepuri NB, Pan T, Hatzoglou M. Angiogenin-cleaved tRNA halves interact with cytochrome c, protecting cells from apoptosis during osmotic stress. Mol Cell Biol 2014; 34(13): 2450–2463

[40] Kumar P, Anaya J, Mudunuri SB, Dutta A. Meta-analysis of tRNA derived RNA fragments reveals that they are evolutionarily conserved and associate with AGO proteins to recognize specific RNA targets. BMC Biol 2014; 12(1): 78

[41] Kumar P, Kuscu C, Dutta A. Biogenesis and function of transfer RNA-related fragments (tRFs). Trends Biochem Sci 2016; 41(8): 679–689

[42] Guan L, Karaiskos S, Grigoriev A. Inferring targeting modes of Argonaute-loaded tRNA fragments. RNA Biol 2020; 17(8): 1070–1080

[43] Venkatesh T, Suresh PS, Tsutsumi R. tRFs: miRNAs in disguise. Gene 2016; 579(2): 133–138

[44] Lyons SM, Kharel P, Akiyama Y, Ojha S, Dave D, Tsvetkov V, Merrick W, Ivanov P, Anderson P. eIF4G has intrinsic G-quadruplex binding activity that is required for tiRNA function. Nucleic Acids Res 2020; 48(11): 6223–6233

[45] Ivanov P, O’Day E, Emara MM, Wagner G, Lieberman J, Anderson P. G-quadruplex structures contribute to the neuroprotective effects of angiogenin-induced tRNA fragments. Proc Natl Acad Sci USA 2014; 111(51): 18201–18206

[46] Lyons SM, Gudanis D, Coyne SM, Gdaniec Z, Ivanov P. Identification of functional tetramolecular RNA G-quadruplexes derived from transfer RNAs. Nat Commun 2017; 8(1): 1127

[47] Nguyen TM, Kabotyanski EB, Dou Y, Reineke LC, Zhang P, Zhang XHF, Malovannaya A, Jung SY, Mo Q, Roarty KP, Chen Y, Zhang B, Neilson JR, Lloyd RE, Perou CM, Ellis MJ, Rosen JM. FGFR1-activated translation of WNT pathway components with structured 5′ UTRs is vulnerable to inhibition of EIF4A-dependent translation initiation. Cancer Res 2018; 78(15): 4229–4240

[48] Raza F, Waldron JA, Quesne JL. Translational dysregulation in cancer: eIF4A isoforms and sequence determinants of eIF4A dependence. Biochem Soc Trans 2015; 43(6): 1227–1233

[49] Biffi G, Di Antonio M, Tannahill D, Balasubramanian S. Visualization and selective chemical targeting of RNA G-quadruplex structures in the cytoplasm of human cells. Nat Chem 2014; 6(1): 75–80

[50] Kwok CK, Merrick CJ. G-Quadruplexes: prediction, characterization, and biological application. Trends Biotechnol 2017; 35(10): 997–1013

[51] Zhan S, Yang P, Zhou S, Xu Y, Xu R, Liang G, Zhang C, Chen X, Yang L, Jin F, Wang Y. Serum mitochondrial tsRNA serves as a novel biomarker for hepatocarcinoma diagnosis. Front Med 2022; 16(2): 216–226

[52] Zhu L, Li J, Gong Y, Wu Q, Tan S, Sun D, Xu X, Zuo Y, Zhao Y, Wei YQ, Wei XW, Peng Y. Exosomal tRNA-derived small RNA as a promising biomarker for cancer diagnosis. Mol Cancer 2019; 18(1): 74

[53] Mo D, Jiang P, Yang Y, Mao X, Tan X, Tang X, Wei D, Li B, Wang X, Tang L, Yan F. A tRNA fragment, 5′-tiRNA^Val, suppresses the Wnt/β-catenin signaling pathway by targeting FZD3 in breast cancer. Cancer Lett 2019; 457: 60–73

[54] Wu Y, Yang X, Jiang G, Zhang H, Ge L, Chen F, Li J, Liu H, Wang H. 5′-tRF-GlyGCC: a tRNA-derived small RNA as a novel biomarker for colorectal cancer diagnosis. Genome Med 2021; 13(1): 20

[55] Yu M, Lu B, Zhang J, Ding J, Liu P, Lu Y. tRNA-derived RNA fragments in cancer: current status and future perspectives. J Hematol Oncol 2020; 13(1): 121

RIGHTS & PERMISSIONS

Higher Education Press