tRNA-derived fragment AS-tDR-008272 promotes NSCLC tumorigenesis by propelling glycolytic metabolism via regulating HJURP

Highlights

• •AS-tDR-008,272 is a novel oncogenic tRF (tRNA-derived fragment) in NSCLC.
• •AS-tDR-008,272 promotes NSCLC malignancy by enhancing glycolytic metabolism of NSCLC cells.
• •AS-tDR-008,272 regulates NSCLC cells by targeting HJURP through activating its translation and transcription.
• •AS-tDR-008,272-HJURP axis may be a potential therapeutic target of NSCLC.

Abstract

Background

Increasing reports have highlighted the roles of tRNA-derived fragments (tRFs) in the pathological processes of cancers. However, the biogenesis, precise functions, and underlying mechanisms of tRFs in non-small-cell lung cancer (NSCLC) remain largely unexplored.

Methods

Dysregulated tRFs in NSCLC were identified by tRF & tiRNA sequencing, qRT-PCR, and fluorescence in situ hybridization (FISH) assays. The biological roles of tRF in cell malignant phenotypes and aerobic glycolysis were determined by functional experiments in vitro and in vivo. The clinical significance of tRF was assessed in human NSCLC samples using FISH and ¹⁸F-FDG PET/CT scan. The underlying mechanisms of tRF were explored by RNA sequencing, dual-luciferase assay, polysome profiling, Western blotting, qRT-PCR, and rescue experiments.

Results

We identified a novel tRF, AS-tDR-008,272, which was upregulated in NSCLC tumor tissues, plasma, and in NSCLC cells. We revealed that the biogenesis of AS-tDR-008,272 was induced by angiogenin (ANG) under hypoxic conditions. Functional assays indicated that AS-tDR-008,272 promoted proliferation, migration, and glycolysis metabolism capacities of NSCLC cells. While inhibition of AS-tDR-008,272 suppressed NSCLC cell growth in vitro and in vivo. Clinically, higher expression of AS-tDR-008,272 in NSCLC tumor tissues was associated with advanced TNM stage, poorer survival, and SUVmax of ¹⁸F-FDG PET/CT scan in NSCLC patients. Mechanistically, AS-tDR-008,272 enhanced NSCLC cell malignancy by interacting with HJURP through a dual regulatory mechanism: binding to the coding sequence (CDS) of HJURP mRNA to regulate its transcription and enhancing the abundance of HJURP mRNA in polysomes to promote its translation.

Conclusions

Our findings reveal a previously unrecognized tRF-mediated glycolysis reprogramming mechanism in NSCLC tumorigenesis. AS-tDR-008,272 may represent a promising prognostic marker and therapeutic target for NSCLC.

Graphical abstract

Background

Lung cancer is the leading cause of cancer-related deaths and an important health care burden in the world [1]. According to the global cancer statistics data, there were 2.48 million new cases and 1.82 million deaths of lung cancer in 2022 (https://gco.iarc.fr/). Among patients with lung cancer, approximately 85% of patients are non-small-cell lung cancer (NSCLC) subtype [2]. In recent years, although tremendous progress has been made in diagnosis, targeted treatment, and immunotherapy, the 5-year overall survival rate for NSCLC remains dismal [3,4]. Therefore, novel diagnostic biomarkers and therapeutic targets remain urgently needed to optimize the prognosis and therapeutic effect of this disease.

Transfer RNAs (tRNAs) are the most abundant non-coding RNAs that account for 4–10% of the total RNAs in cells [5]. The primary biological function of tRNA is to decode the triplet codon on mRNAs and transport specified amino acids for protein synthesis. Under stress conditions, tRNAs can be cleaved by specific ribonucleases (e.g., Dicer and ANG) to yield tRNA-derived small RNAs (tsRNAs), including tRNA-derived fragments (tRFs) and tRNA-derived stress-induced RNAs (tiRNAs) [6]. In recent years, increasing reports have shown that tsRNAs are associated with many physiological and pathological processes, including gene expression [7], ribosome genesis [8], translation [9], and epigenetic regulation [10]. In particular, tsRNAs can regulate tumorigenesis at multiple levels. For instance, tRF-21, which is derived from tRNA^GlyGCC, could suppress pancreatic ductal adenocarcinoma progression by targeting the hnRNP L-DDX17 pathway [11]. Hypoxia-induced tRFs may replace oncogenic RNA-binding protein YBX1 from the 3′-UTRs of oncogenic transcripts, leading to its degradation and thus suppressing breast cancer metastasis [12]. LeuCAG3′tsRNA binds to RPS28 and RPS15 to enhance their translation, leading to increased cell proliferation and tumor growth [8]. tRF-3022b can mediate cell apoptosis and M2 macrophage polarization via binding tocytokines in colorectal cancer [13]. Recently, we have reported that AS-tDR-007333 enhanced the malignancy of NSCLC through HSPB1-mediated histone modification and ELK4-induced MED29 transcription [14]. In summary, these findings indicate a critical role of tRFs in cancer development and progression. Nevertheless, the biogenesis, biological functions, and the molecular mechanisms of most tRFs in NSCLC remain largely unknown.

In this study, we identified a novel tRF termed AS-tDR-008272, which was induced by hypoxia-activated ANG and was upregulated in patient plasma, tumor tissues, and in NSCLC cells. We demonstrated that AS-tDR-008272 promoted the malignant phenotypes and glycolysis metabolism of NSCLC cells. We further confirmed that AS-tDR-008272 expression level was positively correlated with advanced stage, poor prognosis, and higher SUVmax of PET-CT scan in patients with NSCLC. Mechanistically, AS-tDR-008272 drove the malignancy and glycolysis of NSCLC cells by enhancing the translation and transcription of HJURP. Therapeutically, AS-tDR-008272 inhibitor significantly suppressed NSCLC tumor growth in vivo. Collectively, our findings reveal that AS-tDR-008272 is a hypoxia-responsive oncogenic tRF and might serve as a promising therapeutic target and biomarker for NSCLC.

Methods

Patients and clinical samples

Pre- and post-operation blood samples were collected from NSCLC patients who underwent surgical resection in the Department of Thoracic Surgery at Shenzhen People’s Hospital, China. All patients were diagnosed clinically and pathologically with NSCLC (Table S1.). None of the patients had received radiotherapy or chemotherapy before surgery. Written informed consent was obtained from each study patient. The study was performed in accordance with the Declaration of Helsinki and approved by the Medical Ethics Committee of Shenzhen University Health Science Center (Approved no 2016,002).

tRF & tiRNA sequencing

tRF & tiRNA sequencing was performed as described previously [14]. Briefly, total RNA was extracted from plasma using the TRIzol® reagent (Invitrogen, Shanghai, China). To remove RNA modifications that may disturb small RNA library construction, the extracted RNA was pretreated by 3′-aminoacyl deacylation, 2′,3′-cyclic phosphate, and 5′-OH phosphorylation. The libraries were prepared using the NEBNext® Multiplex Small RNA Library Prep Set for Illumina® kit (New England BioLabs, MA, USA). The sequencing analysis was performed on the Illumina NextSeq 500 system at Aksomics Inc. (Shanghai, China). Sequencing reads were aligned to the tRNA sequences database at GtRNAdb using the bowtie software. The tRF expression levels were normalized to the number of transcripts per million of total aligned tRNA reads (TPM).

Cell cultures

Two human NSCLC cell line A549 (cat#0510–07), H358 (cat#ORC0260) and a normal human lung bronchial epithelial cell (BEAS-2B, cat#GNHu27) were obtained from the Cell Bank of Chinese Academy of Biological Sciences (Shanghai, China). Short tandem repeat (STR) analysis was performed to verify the authentication of cells and routine screening was conducted to exclude mycoplasma contamination. Cells were cultured under standard conditions with DMEM or RPMI-1640 medium (Gibco, Shanghai, China) containing 10% (v/v) fetal bovine serum (FBS) (Gibco, Shanghai, China) and 1% (v/v) penicillin-streptomycin (P/S). Cells were incubated in a humidified atmosphere of 5% CO₂ at 37°C and each cell culture analysis was conducted in triplicate.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNAs were isolated from cells using the TRIzol® reagent (Invitrogen, Shanghai, China) and reversely transcribed with PrimeScript RT reagent Kit with gDNA Eraser (Takara Bio, Japan). qRT-PCR was conducted in a qTOWER3 PCR detecting system (Analytik Jena AG, Germany), using the TB Green Premix Ex Taq kit (Takara Bio, Japan). Expression levels of gene or tRF were normalized to that of β-actin and U6, respectively, using the 2^−ΔΔCt method. Each qRT-PCR assay was repeated three times. Primer sequences are listed in Table S 4.

Cell transfection

The AS-tDR-008272 mimic, AS-tDR-008272-inhibitor, and their corresponding normal controls (NCs) were designed and synthesized by Ribobio Co. (Guangzhou, China). The HJURP over-expression plasmids and si-HJURP vector were constructed by Gene Pharma (Shanghai, China). These plasmids were transfected into cells by the Lipofectamine 3000 reagent (Invitrogen, Shanghai, China). Sequencing information for interfering oligos is presented in Table S4.

Cell viability assay

Cell viability was analyzed by the cell counting kit-8 (CCK-8) kit (Dojindo, Shanghai, China) according to the manufacturer’s protocol. Briefly, cells (5000 per well) were seeded in 96-well plates and allowed to adhere overnight. After cultured at 0, 24, 48, 72, and 96 h time points, 10 μl CCK-8 reagent was added to each well and incubated at 37°C for 1 h. Cell viability was determined by detecting the optional density (OD) value at 450 nm, using the FLx800 Fluorescence Microplate Reader (BioTek, Shanghai, China).

Transwell migration assay

Transwell migration assay was performed as previously described [14], using the transwell chambers (BD Falcon, USA). Briefly, cells (1 × 10⁵) suspended in 500 μl serum-free medium were seeded in the upper part of the chambers, and 600 μl of RPMI-1640 medium with 10% FBS was added into the lower part of the chambers. Cells were allowed to migrate for 24 h and then fixed with methyl alcohol and stained by 0.1% crystal violet solution. Chamber images were photographed using an inverted microscope (Olympus, Tokyo, Japan) at ×100 magnification and the migrated cells were counted by ImageJ software.

Wound healing assay

Cell invasion capacity was detected by wound healing assay. In brief, cells were seeded in the 6-well plate and the cell monolayer was scraped using a 10 μl pipette tip to make a wound and allowed the wound to heal for 24 h.The wound healing rates were calculated by comparing the initial width at 0 h with the residual gap at 24 h (the width of wound at 0 h–the width of wound at 24 h)/the width of wound at 0 h.

Flow cytometry

Apoptotic rate of cells was detected using the Annexin V FITC Apoptosis Detection Kit (Dojindo, Shanghai, China) with the Annexin V-FITC and propidium iodide (PI) as fluorescent dyes. CytoFlex flow cytometer (Beckman Coulter, Shanghai, China) equipped with CytExpert software (version 2.3) was used to define the apoptotic type of cells.

Immunohistochemistry (IHC) staining

The IHC staining was examined using the IHC kit (Santa Cruz, Shanghai, China). Briefly, tissue samples were paraformaldehyde-fixed, paraffin-embedded, and sectioned. The slides were deparaffinized and dehydrated by xylene and alcohol, then incubated with the primary antibodies (anti-Ki-67, anti-HJURP) at 4°C overnight, followed with horseradish peroxidase (HPR)-labelled secondary antibody at 37 C for 1 h. Lastly, samples were stained with DAB (3,3-diaminobenzidine) and hematoxylin. The intensity of staining was observed under microscope (Olympus, Japan) with Aperio ImageScope software.

Tissue microarray analysis

A NSCLC tissue microarray (HLugA180Su04) including 92 paired NSCLC tumor tissues and adjacent normal tissues was purchased from Outdo Biotech (Shanghai, China). Alkaline phosphatase-labeled AS-tDR-008272 probe was designed and synthesized by Bosterbio (Wuhan, China). Fluorescence in situ hybridization (FISH) assay was performed to measure the probe signals using the Fluorescent in Situ Hybridization Kit (Bosterbio, Wuhan, China). Images were taken by a fluorescence microscope (Nikon, Tokyo, Japan). The positive stained rate of cells, the intensity of staining, and the staining index (SI) were scored and calculated using the methods as described previously [14]. Median SI values ≥3 was defined as high expression and SI <3 was classified as low expression.

Quantification of lactate and pyruvate levels

Lactate and pyruvate concentrations in the supernatant of cell culture medium were quantified using the lactate assay kit and pyruvate assay kit (Jiancheng Bioengineering, Nanjing, China). Synergy HTX Multi-Mode Microplate Reader (BioTek, Guangzhou, China) was used to detect L-lactate at 530 nm and pyruvate at 505 nm, respectively. The levels of pyruvate and lactate were calculated by calibrating against the standard curves.

The extracellular acidification rate (ECAR) analyses

ECAR assay was conducted with the XF Glycolysis Stress Test Kit (Agilent, Beijing, China), following the manufacturer's protocols. Briefly, glucose, oligomycin (an oxidative phosphorylation inhibitor), and 2-deoxyglucose (2-DG, a glycolytic inhibitor) were sequentially incubated with cells at predefined time points in a Seahorse XF96 cell culture microplate. ECAR was measured and calculated by the Agilent Seahorse XFe96 Analyzer (Agilent Technologies, CA, USA), Each sample was examined in three independent replicates.

Animal experiments

4-week-old BALB/c nude mice were purchased from Shanghai Laboratory Animal Center (Shanghai, China). Agomir AS-tDR-008272-inhibitor and its NC were synthesized and labeled with cy5 by Ribobio Co. (Guangzhou, China). A549 cells (6 × 10⁶) were subcutaneously injected into the right upper back of the nude mice, and the growth of xenografts was monitored by IVIS Lumina K II in vivo imaging system (PerkinElmer, Shanghai, China). After the establishment of xenograft (tumor volume≥100 mm³), mice were randomly divided into three groups (six in each group): tRF inhibitor group, Inhibitor-NC group, and blank control group. Then 5 nmol tRF-inhibitor, inhibitor-NC, or equal volume of PBS were injected into the tumor mass once every 3 days, respectively, for a total of five injections (Fig. 5A). The tumor volumes were monitored and calculated by the equation V = (length x width²)/2 once every 3 days. Animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Shenzhen University Medical School (approved No IACUC-202,300,187).

Fig. 5.

Silencing of AS-tDR-008272 suppresses NSCLC tumor grow in vivo. (A) Schematic overview and timeframe of animal experiments. (B) Representative bioluminescence images of the xenograft tumors receiving inhibitor or inhibitor-NC treatment. (C) Growth curves of xenograft tumors indicating that the knockdown of AS-tDR-008272 inhibits tumor progression. (D) The mean weight of xenograft tumors in mice after different treatments. (E) Image of xenograft tumors resected from nude mice-bearing A549 cells with or without AS-tDR-008272-inhibitor treatment. (F) Representative images of IHC staining of Ki67 and HJURP protein from the tumor sections of indicated experimental groups. (G) Dynamic body weight changes of mice during the experiment. *P < 0.05, **P < 0.01, ***P < 0.001.

RNA-sequencing (RNA-seq)

RNA-seq analysis was performed as previously described [14]. In brief, total RNA from A549 cells was isolated using the TRIzol reagent and the ribosomal RNA (rRNA) was removed by the ribosomal RNA removal kit. Sequencing library was constructed using the KAPA Stranded RNA-Seq Library Prep Kit (Illumina) and the cDNA products were enriched by PCR reactions. RNA-seq was performed on an Illumina NovaSeq 6000 platform at Aksomics Inc. (Shanghai, China) with 150-bp paired end reads, and the raw data have been deposited in the CNSA database with accession number CNP0007692.

Western blotting

Western blotting (WB) was conducted as previously reported [14]. Briefly, cell extracts were obtained using the RIPA lysate (Solarbio, Beijing, China). Proteins were separated by 10% SDS-PAGE (sodium dodecyl sulfate–polyacrylamide gel electrophoresis) and were blotted onto polyvinylidene fluoride (PVDF) membranes. Following blocking with the stop buffer (Beyotime, Shanghai, China) at room temperature for 20 min, the membranes were incubated overnight at 4°C with primary antibodies. Then the membranes were incubated with the secondary antibody (Proteintech, Wuhan, China) for 2 h. Signals were captured with ImageJ system and normalized to those of β-actin. Information for antibodies is provided in Table S5.

Dual‑luciferase reporter assay

Luciferase reporter assays were performed by co-transfecting 20 nM AS-tDR-008272 mimic or NC oligonucleotides with firefly luciferase reporter (100 ng) and Renilla plasmid (100 ng) into HEK293 or A549 cells, respectively, using the Lipo2000 reagent (Invitrogen, Shanghai, China). 48 h after transfection, cells were collected and firefly luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega, Shanghai, China) and normalized to the Renilla luciferase activity.

Polysome profiling

Polysome profiling was conducted as described previously [15]. In brief, A549 cells (1 × 10⁷) were pre-treated with cycloheximide (100 μg/ml) for 10 min. Cell lysates were isolated by polysome lysis buffer (PEB; 20 mM Tris–HCl, pH 7.5, 50 mM KCl; 10 mM MgCl2; 1 mM DTT; 100 μg/ml CHX; 200 μg/ml Heparin). Polysomes were separated on a 5–50% sucrose gradient buffer. Sucrose fractions (n = 13) of equal volume were collected using the BioComp Gradient Station (Fredericton, Canada). Then, fraction 1–6 were pooled together as the lightweight fraction (monosomes), and fraction 7–13 were combined together as the heavy fraction (polysomes) [15]. TRIzol reagent (Life Technologies, Shanghai) was used to extract total RNA from monosomes and polysomes, respectively. mRNAs were reverse transcribed and amplified by qPCR assay.

In silico analysis

Web servers (http://gepia.cancer-pku.cn/, https://kmplot.com/, https://ualcan.path.uab.edu/) were used to retrieve cancer OMICS data (TCGA, MET500, CPTAC and CBTTC) and to compare the expression levels of HJURP protein and gene between NSCLC tumor tissues and adjacent normal tissues, and the association between HJURP expression level and patient outcomes.

Statistical analysis

Each in vitro assay was performed at least three times. Experimental data were presented as mean ± standard deviation (SD). All statistical analyses were conducted using SPSS 26.0 (IBM, Chicago, USA) and GraphPad Prism (v.8). The Student’s T–test or the Mann–Whitney U test was used to evaluate the statistical significance between groups. Paired-t-test was applied to compare the difference of AS-tDR-008272 expression level between tumor tissues and paired normal tissues. The Kaplan–Meier method, log-rank test, and Cox regression model were used to assess the survival curves of NSCLC patients. Correlations were measured via Spearman correlation analysis. For all analyses, a P-value <0.05 was considered statistically significant (*P < 0.05, **P < 0.01, and ***P < 0.001).

Results

AS-tDR-008272 is elevated in NSCLC and associated with patient prognosis

A total of nine pairs of pre- and post-surgery plasma samples were used to identify differentially expressed (DE) tRFs in NSCLC patients, using the tsRNA sequencing analysis [15]. The results showed that tsRNA expression profiles in pre-operation plasma were distinct from that of post-operation plasma samples (Fig. 1A), suggesting that DE tRFs reflected the existence of tumor tissues. In particular, 4 up-regulated (fold change ≥2.0, P < 0.05) tRFs in pre-operation samples were identified, with AS-tDR-008272 (derived from pre-Ser-TGA-1–1) ranked as the top DE tRF (fold change = 4.02, P = 0.0194) (Fig.1B–C). To determine whether AS-tDR-008272 was upregulated endogenously in NSCLC cells, AS-tDR-008272 expression levels in NSCLC cells (A549, H358) were compared to that of BEAS-2B cells. The results indicated that AS-tDR-008272 expression levels in NSCLC cells were significantly higher than that in BEAS-2B cells (Fig.1D).

Fig. 1.

AS-tDR-008,272 is upregulated in NSCLC and its overexpression is associated with poor prognosis. (A) Venn diagram analysis reveals distinct differentially expressed tRFs in pre- and post-operation plasma samples. (B, C) Among the most upregulated tRFs in pre-operation plasma, AS-tDR-008272 displayed the highest fold change and smallest P-value. (D) The expression levels of AS-tDR-008272 in NSCLC cells were higher than that in normal bronchial epithelial cells (BEAS-2B). (E) The expression of AS-tDR-008272 in NSCLC tumor tissues and matched adjacent normal tissues verified by FISH assays in tissue microarrays. AS-tDR-008272 expression level in NSCLC tumor tissues was significantly higher than that in the adjacent tissues and increased as NSCLC progressed to more advanced stages. (G) Representative FISH images of AS-tDR-008272 expression in NSCLC tumor tissues and in different TNM stages. (F) Higher AS-tDR-008272 levels were associated with shorter overall survival in NSCLC patients (Log-rank P value = 0.008; cut-off value: score = 3.0). *P < 0.05, **P < 0.01, ***P < 0.001.

To investigate the clinical significance of AS-tDR-008272 expression level in NSCLC, AS-tDR-008272 expression levels were analyzed on a tissue microarray containing NSCLC tumor tissues and adjacent tissues. FISH assay revealed that AS-tDR-008272 expression levels in NSCLC tumor tissues were significantly higher than that in adjacent tissues. Notably, expression levels of AS-tDR-008272 in advanced TNM stages were significantly higher than that in early stages (P = 0.0011) (Fig. 1E, G), indicating that AS-tDR-008,272 expression level was associated with the clinical severity of NSCLC. In addition, Kaplan-Meier survival analysis indicated that patients with higher expression levels of AS-tDR-008,272 had a shorter overall survival rate than those with low expression of AS-tDR-008,272 (Fig. 1F). Moreover, multivariate Cox proportional hazard analysis showed that higher expression level of AS-tDR-008272 in NSCLC tumor tissues was an independent predictor for worse prognosis (HR = 2.09, P = 0.018). Collectively, these data indicate that AS-tDR-008272 is overexpressed in NSCLC and positively associated with poor NSCLC prognosis.

Biogenesis of AS-tDR-008272 is induced by angiogenin (ANG) under hypoxia conditions

Having identified that AS-tDR-008272 was highly expressed in NSCLC, we next investigated the molecular mechanism leading to AS-tDR-008272 upregulation. Since previous studies have suggested that hypoxia is a hallmark of solid tumors [19] and hypoxia could affect the expression of tRF [20], we hypothesized that hypoxia might drive AS-tDR-008272 production in NSCLC cells. To test this hypothesis, we cultured BEAS-2B, A549 and H358 cells under hypoxia or normoxia conditions, respectively, and then examined the expression levels AS-tDR-008272 in these cells using qPCR assay. The results showed that AS-tDR-008272 expression levels were substantially higher under hypoxic conditions than that in normoxia for all of these cells (Fig. 2A), suggesting that hypoxia activates AS-tDR-008272 biogenesis in both normal and NSCLC cells.

Fig. 2.

Hypoxia activates ANG expression to induce AS-tDR-008272 over-production. (A) AS-tDR-008272 is upregulated under hypoxic conditions (1% O₂) in various cell lines. (B) Hypoxia induces upregulation of ANG in BEAS-2B and NSCLC cells. (C) Hypoxia promotes the expression of Dicer gene in BEAS-2B and A549 cells, but not in H358 cell line. (D) Hypoxia promotes ANG protein expression levels in both A549 and H358 cells. (E) ANG expression is upregulated under hypoxia conditions in NSCLC cells. (F) Overexpression of ANG enhances AS-tDR-008272 expression in A549 cells. (G) Upregulation of ANG promotes AS-tDR-008272 expression in H358 cells. (H) Silencing of ANG suppresses AS-tDR-008272 expression in A549 cells. (I) Knockdown of ANG decreases AS-tDR-008272 expression in H358 cells. *P < 0.05, **P < 0.01, ***P < 0.001.

Given that most tRFs are cleaved from tRNAs by ribonuclease (RNase) [6], we sought to examine which RNase was responsible for AS-tDR-008272 biogenesis. We incubated A549 and H358 cells under hypoxia and normoxia conditions, respectively. Then we compared the expression levels of Dicer and ANG, two major ribonucleases for tRF biogenesis [9], between normoxia- and hypoxia-cultured cells. qRT-PCR analyses revealed that the expression levels of ANG and Dicer in both hypoxia-treated NSCLC cells were significantly higher than that in normoxia-cultured cells (Fig.2B–C). With ANG displayed stronger response to hypoxia. Further analyses showed that hypoxia promoted the expression of ANG protein and gene in both A549 and H358 cells (Fig. 2D, 2E). These results indicated that ANG was a hypoxia responsive RNase in NSCLC cells. To investigate whether ANG may play a role in AS-tDR-008272 biogenesis, we constructed ANG overexpression plasmids and si-ANG to overexpress or knockdown ANG. Intriguingly, transfection of ANG-overexpressed plasmids into NSCLC cells led to a higher expression of AS-tDR-008272 (Fig. 2F, G). In contrast, knockdown of ANG resulted in a decreased expression level of AS-tDR-008272 in NSCLC cells (Fig. 2H, 2I). Collectively, these data indicate that hypoxia-induced AS-tDR-008,272 biogenesis is mediated by the upregulation of ANG in NSCLC cells.

AS-tDR-008272 promotes proliferation, migration, and invasion capacities of NSCLC cells

To examine the oncogenic effects of AS-tDR-008272 on NSCLC cells, we transfected AS-tDR-008272 mimic and its inhibitor into NSCLC cells, respectively (Fig. S1). CCK-8 assays showed that overexpression of AS-tDR-008272 markedly enhanced the proliferation rates of NSCLC cells (Fig. 3A, B), whereas knockdown of AS-tDR-008272 significantly suppressed cell viability (Fig. 3C, D). In agreement with this, up-regulation of AS-tDR-008272 also enhanced the migratory and invasive capabilities of NSCLC cells by wound healing and transwell assays (Fig. 3E, F, I, J), while down-regulation of AS-tDR-008272 resulted in opposite effects (Fig. 3G, H, K, L). Nevertheless, the apoptosis assay suggested that AS-tDR-008272 expression had limited influence on the apoptosis rate of NSCLC cells (Fig. S2). Taken together, these results indicate that AS-tDR-008272 plays an oncogenic role in the NSCLC cells.

Fig. 3.

AS-tDR-008272 facilitates proliferation, migration, and invasion of NSCLC cells. (A) AS-tDR-008272 overexpression increases the proliferation rate of A549 cells. (B) Upregulation of AS-tDR-008272 promotes H358 cell proliferation. (C) Knockdown of AS-tDR-008272 decreases the proliferation rate of A549 cells. (D) Downregulation of AS-tDR-008272 inhibits H358 cell proliferation. (E) Higher expression of AS-tDR-008272 enhances migration capacity of A549 cells. (F) Overexpression of AS-tDR-008272 promotes migration rate of H358 cells. (G) Lower expression of AS-tDR-008272 suppresses migration ability of A549 cells. (H) Downregulation of AS-tDR-008272 inhibits migration capacity of H358 cells. (I) Upregulation of AS-tDR-008272 facilitates wound healing ability of A549 cells. (J) Overexpression of AS-tDR-008272 increases wound invasion capacity of H358 cells. (K) Silencing of AS-tDR-008272 inhibits wound healing ability of A549 cells. (L) Inhibition of AS-tDR-008272 reduces the invasion capacity of H358 cells. *P < 0.05, **P < 0.01, ***P < 0.001.

AS-tDR-008272 enhances glycolysis metabolism in nsclc

Since hypoxia microenvironment is linked to glycolysis metabolism [16,17] and that glycolysis is the dominant metabolic process in solid tumors [18], we speculated that hypoxia-induced AS-tDR-008272 may be involved in glycolysis metabolism in NSCLC cells. To test this hypothesis, we assessed the effect of AS-tDR-008272 expression level on the production of lactate and pyruvate, key end products of glycolysis metabolism in cells. The pyruvate assay demonstrated that overexpression of AS-tDR-008272 promoted pyruvate production in both A549 and H358 cells (Fig. 4A–B). Similarly, enforced upregulation of AS-tDR-008272 increased the production of lactate from NSCLC cells (Fig. 4C-D). Whereas the production levels of both lactate and pyruvate were significantly decreased in AS-tDR-008272-knockdown cells (Fig. 4E–H). Moreover, ECAR analysis revealed that AS-tDR-008272 overexpression promoted the glycolytic capacity in both A549 and H358 cells. (Fig. 4I-J). In contrast, silencing of AS-tDR-008272 suppressed the ECAR levels in NSCLC cells (Fig. 4K-L). To evaluate the clinical relevance of AS-tDR-008272 in NSCLC glucose metabolism, we analyzed the correlation between SUVmax (maximum standardized uptake value, a surrogate for aerobic glycolysis) of PET-CT scan and the expression levels of AS-tDR-008272 in tumor tissues from 19 NSCLC patients. FISH assay showed that the expression levels of AS-tDR-008272 in tumors were positively correlated with the SUVmax values (R = 0.983, P < 0.001) (Fig. 4M). As indicated by the ¹⁸F-FDG PET/CT imaging, tumors with lower AS-tDR-008272 expression (lower panel) had a relatively weak SUVmax signal, while ¹⁸F-FDG uptake in AS-tDR-008272 upregulation tumor was much higher (upper panel) (Fig. 4N). Collectively, these data suggest that AS-tDR-008272 accelerates the glycolysis metabolism in NSCLC tumor tissues.

Fig. 4.

AS-tDR-008272 facilitates glycolysis metabolism of NSCLC in vitro and in vivo. (A) Overexpression of AS-tDR-008272 promotes pyruvate production from A549 cells. (B) Upregulation of AS-tDR-008272 enhances pyruvate generation from H358 cells. (C) Higher expression of AS-tDR-008272 increases lactate production from A549 cells. (D) Transfection of AS-tDR-008272 mimic promotes the production of lactate in H358 cells. (E) Low expression of AS-tDR-008272 suppresses pyruvate generation from A549 cells. (F) Downregulation of AS-tDR-008272 reduces pyruvate production from H358 cells. (G) Inhibition of AS-tDR-008272 decreases lactate production from A549 cells. (H) Low expression of AS-tDR-008272 represses lactate production from H358 cells. (I) Upregulation of AS-tDR-008272 promotes ECAR level in A549 cells. (J) AS-tDR-008272 enhances ECAR level in H358 cells. (K) Knockdown of AS-tDR-008272 suppresses ECAR levels in A549 cells. (L) Inhibition of AS-tDR-008272 downregulates ECAR levels in H358 cells. (M) The AS-tDR-008272 FISH staining scores were positively correlated with the SUVmax of PET-CT scan in patients with NSCLC. (N) Representative ¹⁸F-FDG PET/CT images in NSCLC patient with low level of AS-tDR-008272 (FISH score<1) and SUVmax (2.5) vs. that of high level of AS-tDR-008272 (FISH score=4.5) and SUVmax (12.3). Data are presented as the means ± SEMs. ** P < 0.01 and *** P < 0.001.

AS-tDR-008272 downregulation suppresses nsclc tumor growth in vivo

Given that silencing of AS-tDR-008272 decreased the proliferation rate of NSCLC cells, we hypothesized that AS-tDR-008272 inhibition could reduce tumor growth in vivo. By treating NSCLC xenograft tumor in nude mice with PBS, NC, and AS-tDR-008272-inhibitor respectively (Fig. 5A–B), we observed that knockdown of AS-tDR-008,272 significantly decreased NSCLC tumor weight and tumor volume in nude mice (Fig. 5C–E). But the effect of AS-tDR-008272-inhibitor on mice bodyweight was not statistically significant, suggesting that the toxicity of AS-tDR-008272-inhibitor on mice was relatively limited (Fig. 5G). IHC assay also showed that AS-tDR-008272 downregulation diminished the expression of HJURP protein, and Ki-67 protein, a biomarker of cell proliferation, in xenograft tumors (Fig. 5F). Thus, our data indicate that AS-tDR-008272 downregulation can effectively decrease NSCLC tumor growth, strongly supporting that AS-tDR-008272 is crucial in the regulation of NSCLC progression.

AS-tDR-008272 regulates NSCLC malignancy and glycolysis by interacting with HJURP

To explore the downstream target gene of AS-tDR-008272 in NSCLC, RNA-seq was performed in AS-tDR-008272 overexpressing NSCLC cells and control cells. The results revealed that AS-tDR-008272 regulated the expression of a number of functional genes (<0.5-fold or >2-fold, P < 0.05) (Fig. 6A, B). Gene Ontology (GO) analysis showed that AS-tDR-008272-regulated genes were mainly enriched in the biological processes involving metabolic and catabolic processes, regulation of gene expression and cell cycle (Fig. 6C, D). GSEA analysis suggested that starch metabolism pathway and P53 pathway, two important pathways in glycolysis metabolism, were significantly enriched in AS-tDR-008272-overexpressing cells (Fig. 6E, F), implying that AS-tDR-008272-dysregulated genes may be implicated in glycolysis metabolism in NSCLC cells. Among these dysregulated genes, HJURP was the top outlier gene with 2.28-fold up-regulation (P = 0.0006) (Fig. 6G). Up-regulation of HJURP by AS-tDR-008272 overexpression was validated by qRT-PCR in NSCLC cells (Fig. 6H). Conversely, AS-tDR-008272 knockdown in NSCLC cells downregulated HJURP expression in both A549 and H358 cells (Fig. 6I). Further Western blot assay confirmed that AS-tDR-008272-inhibitor suppressed HJURP protein expression while upregulation of AS-tDR-008272 enhanced HJURP protein expression in A549 cells (Fig. 6J). Moreover, HJURP was also reported to be upregulated in NSCLC and other types of cancers and associated with poorer prognosis of NSCLC (Fig. S3). These data suggest that AS-tDR-008272 may exert its oncogenic effect on NSCLC through targeting HJURP.

Fig. 6.

HJURP is the target gene of AS-tDR-008272. (A) Volcanoplot analysis indicating differentially expressed genes between AS-tDR-008272-overexpressing A549 cells and control cells. (B) Principal component analysis revealing a clear difference between genetic components of AS-tDR-008272-overexpressing cells and control cells. (C, D) Gene ontology (GO) analysis indicates that AS-tDR-008272-regulated genes are correlated to processes associated with metabolism, cell cycle, and signal transduction. (E, F) Gene Set Enrichment Analysis (GSEA) shows that AS-tDR-008272-regulated genes are enriched in glucose metabolism-related pathways such as starch metabolism and P53 signaling pathways. (G) Heatmap plot showing the differentially regulated mRNAs from RNA-seq data analysis in the AS-tDR-008272-overexpressing A549 cells, indicating that HUJRP is the top up generegulated by AS-tDR-008272. (H-I) qRT-PCR assay validates that overexpression of AS-tDR-008272 promotes HJURP expression while knockdown of AS-tDR-008272 inhibits HJURP expression in NSCLC cells. (J) Knockdown of AS-tDR-008272 inhibits HJURP protein expression but overexpression of AS-tDR-008272 promotes HJURP protein expression in NSCLC cells. *P < 0.05, **P < 0.01, ***P < 0.001.

To determine whether the biological influence of AS-tDR-008272 on NSCLC cells was dependent on HJURP, we conducted a series of rescue experiments by co-transfecting si-HJURP (Fig. S4) and AS-tDR-008272 mimic into NSCLC cells. The EdU assay demonstrated that si-HJURP significantly reversed the promoting effect of AS-tDR-008272 mimic on A549 cell proliferation (Fig. 7A). Results from the lactate assay revealed that si-HJURP also significantly attenuated the effects of AS-tDR-008272 overexpression on the productions of pyruvate and lactate from NSCLC cells (Fig. 7B, C). In addition, the knockdown of HJURP could also diminish the enhancing effects of AS-tDR-008272 on cell proliferation, pyruvate and lactate release from both A549 and H358 cells (Fig. 7D, E). Similar effects were also observed in ECAR assays (Fig. 7G-H). In summary, these data indicate that the oncogenic and pro-glycolytic effects of AS-tDR-008272 on NSCLC cells were mediated by direct interaction with HJURP.

Fig. 7.

AS-tDR-008272 promotes NSCLC cell proliferation and glycolysis by interacting with HJURP. (A) Co-transfection of AS-tDR-008272 mimic and si-HJURP attenuates the effects of AS-tDR-008272 on cell proliferation of A549 cells. (B) si-HJURP partly reverses the effects of AS-tDR-008272 on pyruvate production in A549 cells. (C) Downregulation of HJURP partially eliminates the effects of AS-tDR-008272 overexpression on lactate production in A549 cells. (D) si-HJURP suppresses the effect of AS-tDR-008272 on cell proliferation in H358 cells. (E) si-HJURP abolishes the promotive effect of AS-tDR-008272 on pyruvate generation from H358 cells. (F) si-HJURP diminishes the impact of AS-tDR-008272 on lactate release from H358 cells. (G) Co-transfection of AS-tDR-008272 mimic and si-HJURP decreases the effects of AS-tDR-008272 on ECAR levels in A549 cells. (H) si-HJURP partly reverses the effects of AS-tDR-008272 on ECAR levels in H358 cells. (I) Schematic diagram of the predicted binding site of AS-tDR-008272 on wildtype HJURP mRNA and mutated region of HJURP. (J) The Wildtype (WT) or mutated reporter plasmid was co-transfected into HEK-293T cells with AS-tDR-008272 mimic or negative control. Luciferase activity of WT-HJURP was significantly decreased by AS-tDR-008272 in cells. (K) WT, but not the mutated binding site of HJURP leads to altered luciferase activity by the AS-tDR-008272 mimic in A549 cells. (L) AS-tDR-008272 enhances the abundance of HJURP mRNA in polysomes but decreases the expression levels of HJURP in monosomes. (M) Knockdown of AS-tDR-008272 decreases HJURP mRNA expression levels in polysomes but increases HJURP expression levels in monosomes. (N) Upregulation of AS-tDR-008272 increases the polysome/monosome ratio of HJURP expression levels in A549 cells. (O) Inhibition of AS-tDR-008272 decreases the polysome/monosome ratio of HJURP expression levels in A549 cells. *P < 0.05, **P < 0.01, ****P < 0.0001.

AS-tDR-008272 targets HJURP by binding to the coding sequence (CDS) of HJURP mRNA

As tRFs have been found to be involved in gene and/or protein expression [20,21], we next explored the molecular mechanisms by which AS-tDR-008272 regulated HJURP expression. Using the RNAhybrid tool (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid/), we found that there is one potential binding site for AS-tDR-008272 in the coding sequence (CDS) of HJURP mRNA (Fig. 7I). To elucidate the regulatory role of AS-tDR-008272 on HJURP expression, wild type (WT) or mutant type HJURP sequence fragments containing the potential AS-tDR-008272 binding site were chemically synthesized and subcloned into the downstream of the luciferase gene in the pmirGLO plasmid (Ambion, USA), respectively. Then the wild type and mutant dual-luciferase vectors were co-transfected with AS-tDR-008272 mimic and corresponding NC into HEK293T or A549 cells. Luciferase reporter assays showed that the reporter plasmid with wild type targeting sequence of HJURP caused a significant decrease in luciferase activity in cells transfected with AS-tDR-008272 mimic, whereas reporter plasmid with mu tant sequence of HJURP produced no change in luciferase activity (Fig. 7J-K). Taken together, these data suggest that the AS-tDR-008,272 binding site in the HJURP CDS region is critical for AS-tDR-008,272-mediated HJURP expression.

AS-tDR-008272 facilitates HJURP translation

Because tRF is known to be involved in translation by regulating ribosome activities [8], we further investigated whether AS-tDR-008272 may play a role in HJURP translation. To this end, we transfected AS-tDR-008272 mimic, inhibitor, and their corresponding NCs into A549 cells, respectively. Subsequently we extracted ribosomes from A549 cells and then separated polysomes from monosomes. qPCR assays showed that upregulation of AS-tDR-008272 enhanced the expression levels of HJURP mRNA in polysomes but decreased the abundance of HJURP mRNA in monosomes (Fig. 7L), suggesting that HJURP translation was upregulated. On the contrary, transfection of AS-tDR-008272-inhibitor into A549 cells resulted in a decrease in HJURP mRNA expression level in polysomes but an increase of HJURP mRNA expression in monosomes (Fig. 7M), implying a decrease in translation initiation. Moreover, overexpression of AS-tDR-008272 increased the polysome/monosome ratio of HJURP expression levels (Fig. 7N), and downregulation of AS-tDR-008272 suppressed the polysome/monosome ratio of HJURP expression (Fig. 7O). In summary, these observations indicate that AS-tDR-008,272 promotes HJURP translation by enhancing its abundance in the polysomes of NSCLC cells.

Discussion

Hypoxia is one of the most common microenvironmental features of solid tumors [22], which contributes to cancer progression, recurrence, metastasis, immune suppression, and resistance to chemoradiotherapy [23]. It has been reported that hypoxia-induced non-coding RNAs, including lncRNAs, miRNAs, and circRNAs, are implicated in the development and progression of cancers [[24], [25], [26], [27]]. However, whether tRFs, a new class of small non-coding RNA (sncRNA), may also play a role in hypoxia-mediated malignancy remains elusive. In this study, we identified a novel tRF termed As-tDR-008272 that was induced by ANG under hypoxia conditions. We found that As-tDR-008272 was upregulated in NSCLC and its high expression was associated with poorer prognosis of NSCLC patients. We showed that overexpression of As-tDR-008272 promoted proliferation, migration, and invasion capacities of NSCLC cells. In addition, we revealed that As-tDR-008,272 enhanced glycolysis metabolism in NSCLC cells as well as in tumor tissues of NSCLC patients. Furthermore, we confirmed that inhibition of As-tDR-008272 suppressed NSCLC tumor growth in vivo. Our findings indicate that hypoxia-induced As-tDR-008272 is an oncogenic tRF and a potential therapeutic target of NSCLC.

Although tRFs have been implicated in malignant behaviors of cancers, the exact mechanisms driving the biogenesis of most tRFs are largely unknown. The ANG is a member of the ribonuclease A superfamily that induces tRF production by cleaving the anticodon loop of tRNAs [[28], [29]]. ANG gene has been reported to be upregulated under hypoxia conditions in solid tumors including NSCLC [30,31]. In normal conditions, ANG is inactive in nucleus by binding to ribonuclease inhibitor 1 (RNH1). Under hypoxia or other stress conditions, ANG can translocate to the cytosol where it specifically cleaves tRNAs to generate tsRNAs [30,32]. For example, arsenite-induced overproduction of tRF5-AlaCGC was mediated by ANG in A549 cells [33]. Under hypoxia conditions, ANG induced the production of tRF-Glu-CTC, which in turn triggered the secretion of inflammatory factors [34]. Here, we confirmed that ANG was overexpressed under hypoxia conditions. Moreover, we revealed that upregulation of ANG enhanced the generation of As-tDR-008272 from NSCLC cells. Our data indicates that the biogenesis of As-tDR-008272 is mediated by hypoxia-induced ANG in NSCLC cells.

To date, several hypoxia-related tRFs have been linked to the development and progression of certain types of cancer. It has been found that hypoxia-induced tRF-20-MEJB5Y13 could promote the malignancy of colorectal cancer cells [35]. Under hypoxic conditions, tRNA^Glu-, tRNA^Asp-, tRNA^Gly-, and tRNA^Tyr-derived tRFs may displace YBX1 from the 3′-UTRs of oncogenes, resulting in its degradation and thus suppressing breast cancer metastasis [36]. Hypoxia-induced tRF-20-M0NK5Y93 promoted colon cancer metastasis by binding to MALAT-1 to modulate the expression of SMC1A [37] or by impairing the epithelial-to-mesenchymal transition through targeting Claudin-1 [38]. These findings highlighted the critical roles of hypoxia-induced tRFs in the pathogenesis of cancer. Nevertheless, whether hypoxia-associated tRFs are implicated in the tumorigenesis of NSCLC is unclear. Here, we showed that hypoxia-induced As-tDR-008272 played significant roles in NSCLC malignancy. To our knowledge, this is the first report showing the oncogenic effects of hypoxia-induced tRF in NSCLC.

Reprogramming of energy metabolism is a typical feature of cancer cells. To ensure prompt proliferation, cancer cells rely primarily on rapid but low efficient glycolysis, rather than on the oxidative phosphorylation for energy demands, even in the presence of abundant oxygen [39,40]. Elevated glycolysis can also promote cancer cell malignancy through activating a series of functional pathways, suggesting that targeting glycolysis is a potential therapeutic strategy against cancer [41]. Previously, several types of non-coding RNAs, including lncRNAs, miRNAs, and circRNAs have been reported to be involved in the regulation of glycolysis metabolism in cancers [42]. However, little is known about the potential of tRFs in glycolysis metabolism. In the present study, we found that AS-tDR-008272 promoted glycolysis metabolism in NSCLC cells, as indicated by increased lactate and pyruvate production and enhanced ECAR level. These results were further confirmed in clinical samples in which higher expression of AS-tDR-008272 in tumor tissues was correlated with higher SUVmax of PET-CT scan, a surrogate biomarker of glycolysis metabolism. Our findings reveal a new function of tRF in regulating glycolysis reprograming in NSCLC carcinogenesis.

Holliday junction recognition protein (HJURP) is a histone H3 chaperone that mediates centromeric chromatin assembly, chromosome stability, cell mitosis, and DNA double-strand breaks repair [43]. HJURP may also be involved in the glycolytic signaling pathway in lung adenocarcinoma [44]. Notably, HJURP was highly expressed and correlated with poor prognosis of some malignancies, such as liver [45], breast [46], colorectal [47], and lung cancer [48]. In lung cancer, HJURP protein expression level would increase in NSCLC cells when cells were exposed to DNA-damaging agents such as γ-irradiation and cisplatin [49]. These reports suggested that HJURP might be a critical regulator in cancer tumorigenesis. However, the molecular mechanisms that regulate HJURP expression and the biological functions of HJURP in cancer cells remain largely unclear. Here, we found that HJURP expression was modified by AS-tDR-008272 at transcriptional and translational levels. Moreover, we revealed that HJURP interacted with AS-tDR-008272 to promote the malignancy and glycolysis metabolism of NSCLC cells. We thus elucidate a novel function and mechanism of HJURP in regulation of the malignant phenotypes of NSCLC.

We recognize some limitations of this study. First, while we identified several oncogenic impacts of AS-tDR-008272 on NSCLC, our data was not able to define which is the most significant biological function of AS-tDR-008272. Further studies are needed to characterize the most important biological function of AS-tDR-008272 in NSCLC. Second, our results showed that HJURP was a key target gene of AS-tDR-008272, whether HJURP expression level is associated with NSCLC progression required to be verified in large population of NSCLC cohort. Third, the in vitro experiments were performed in lung adenocarcinoma (LUAD) cells (A549 and H358 cells) only, which is the most common subtype of NSCLC. Results from the present study cannot be simply translated into the biological functions of AS-tDR-008272 in other subtypes of NSCLC. Further analyses in squamous cell carcinoma and large cell carcinoma cell lines may help to better understand the biological functions and regulatory mechanisms of AS-tDR-008272 in NSCLC.

Conclusions

In summary, we identified a hypoxia-driven and ANG-induced AS-tDR-008272 in NSCLC. We confirmed that AS-tDR-008272 was an oncogenic tRF which promoted the malignancy and glycolysis metabolism of NSCLC cells. These effects were achieved by activating HJURP transcription and translation. Our findings not only reveal a previously unrecognized role and mechanism underlying the oncogenic properties of tRF in NSCLC but also present the potential of AS-tDR-008272 as a promising molecular biomarker or therapeutic target for NSCLC.

Declarations

Ethics approval and consent to participate

The study was performed in accordance with the Declaration of Helsinki and approved by the Medical Ethics Committee of Shenzhen University Health Science Center (Approved no 2016,002). Animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Shenzhen University Medical School (approved No IACUC-202,300,187).

Consent for publication

All the authors agreed to be published.

Funding

This work was supported by the National Natural Science Foundation of China (41977372 to RZ), Youth Fund of the National Natural Science Foundation of China (82302977 to WY), Medical Scientific Research Foundation of Guangdong Province (B2025609 to WY), Natural Science Foundation of Guangdong Province (2024A1515012760 to KG), Shenzhen Science and Innovation Commission (JCYJ20240813102013018 to WY, JCYJ20190806154210829 to LY), and the Department of Nanshan District Health and Wellness (NSZD2024067 to LW).

CRediT authorship contribution statement

Duanyang Zhou: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Xiao Li: Writing – original draft, Validation, Resources, Investigation, Formal analysis, Data curation. Yuting Wang: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Lin Yang: Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Data curation. Yongyi Huang: Investigation, Formal analysis. Qianqian Chen: Validation, Software, Investigation, Formal analysis. Yiling Wang: Investigation, Formal analysis, Data curation. Yuxin Wen: Resources, Investigation, Formal analysis. Kaiping Gao: Validation, Project administration, Funding acquisition. Qingchun Liu: Resources, Investigation, Data curation. Xinli Zhou: Validation, Resources, Investigation. Liping Wang: Writing – review & editing, Validation, Funding acquisition, Formal analysis. Wenhan Yang: Writing – review & editing, Validation, Methodology, Investigation, Funding acquisition, Formal analysis. Rihong Zhai: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.