tRNA-derived fragments: mechanism of gene regulation and clinical application in lung cancer

Abstract

Lung cancer, being the most widespread and lethal form of cancer globally, has a high incidence and mortality rate primarily attributed to challenges associated with early detection, extensive metastasis, and frequent recurrence. In the context of lung cancer development, noncoding RNA molecules have a crucial role in governing gene expression and protein synthesis. Specifically, tRNA-derived fragments (tRFs), a subset of noncoding RNAs, exert significant biological influences on cancer progression, encompassing transcription and translation processes as well as epigenetic regulation. This article primarily examines the mechanisms by which tRFs modulate gene expression and contribute to tumorigenesis in lung cancer. Furthermore, we provide a comprehensive overview of the current bioinformatics analysis of tRFs in lung cancer, with the objective of offering a systematic and efficient approach for studying the expression profiling, functional enrichment, and molecular mechanisms of tRFs in this disease. Finally, we discuss the clinical significance and potential avenues for future research on tRFs in lung cancer. This paper presents a comprehensive systematic review of the existing research findings on tRFs in lung cancer, aiming to offer improved biomarkers and drug targets for clinical management of lung cancer.

Introduction

Lung cancer is the leading cause of death among all cancers [1]. It is usually categorized according to histological classification into small cell lung cancer (SCLC), accounting for 15% of cases, and non-small cell lung cancer (NSCLC), accounting for 85%; the latter is divided into adenocarcinoma, squamous cell carcinoma and large cell carcinoma according to location and clinical symptoms [2]. Precision medicine is concerned with finding meticulous and clearer methods of classification or stratification based on the different phenotypes of a disease to develop detailed diagnostic strategies and personalized treatment plans, such as targeted drugs [3]. On the other hand, postchemotherapy targets specific biochemical indicators of cancer, such as biomarkers, to develop novel targeted drugs to improve treatment efficacy and reduce toxic side effects [4]. Postchemotherapy is a practical application of the concept of precision medicine in cancer treatment. With the advent of the era of “precision medicine” and “postchemotherapy”, targeted drugs have become an effective means to improve the clinical efficacy and side effects of treatment for lung cancer, replacing traditional chemotherapy drugs [5, 6]. Although new cancer treatment concepts and strategies have led to significant improvements in clinical lung cancer treatment, the problems of difficult early diagnosis, epithelial mesenchymal metastasis and resistance to therapeutic agents have yet to be resolved [2, 7]. Therefore, it is crucial to further explore the molecular mechanisms regulating the occurrence and development of lung cancer to identify new targets and biomarkers.

Noncoding RNAs (ncRNAs), including microRNAs (miRNAs), circular RNAs (circRNAs) and long noncoding RNAs (lncRNAs), typically do not encode proteins but are involved in the regulation of various cellular physiological processes. Studies have shown that ncRNAs do not operate independently but form a complex network of gene regulation with miRNAs as their core [8]. tRNA-derived fragments, were thought to be random degradation products of tRNAs, ranging in length from 14 to 50 nucleotides (nt). In fact, tRFs derive from tRNA cleaved by RNases at specific modification sites, and this mechanism is strictly and finely regulated. Each type of tRF can only be cleaved by the corresponding RNase and has potential biological value [9, 10]. There is increasing evidence that tRFs regulate intracellular gene transcription, protein translation and epigenetic inheritance. tRFs participate in numerous molecular regulatory mechanisms, such as miRNA-like silencing of target genes, RNA processing and degradation, histone modification, ribosome assembly and activity, and the unfolded protein response [11–14]. In addition, tRFs play a key role in many diseases, such as cancer [15, 16], acute and chronic liver failure [17], COVID-19 [18], and metabolic disorders [19]. In this review, we summarize the classification and modifications of tRFs. Furthermore, the mechanisms by which tRFs regulate gene expression and tumorigenesis in lung cancer are systematically reviewed. Then, bioinformatics studies of tRFs in lung cancer are summarized. Finally, the potential clinical application value of tRFs in lung cancer is discussed.

Classification and modifications of tRFs and tiRNAs

In general, tRFs can be classified into four categories, tRF-1, tRF-3, tRF-5, and i-tRF (tRF-2), according to cleavage site on precursor or mature tRNA, length and sequence. tRF-1s are derived from the cleavage of 5-terminus of precursor tRNAs. Other tRFs are derived from mature tRNAs. In addition, under stress stimulus conditions, the anticodon loop of tRNAs is enzymatically cleaved to form tRNA semimolecules called tiRNAs. tiRNAs are categorized as 5’-tiRNAs and 3’-tiRNAs according to the fragment of the parental tRNAs. The biogenesis of tRFs and tiRNAs involves a variety of endonucleases and undergoes a complex process. Our laboratory has systematically summarized the classification of tRFs and tiRNAs [20]. Building on previous evidence, we identified a number of less common enzymes involved in the cleavage of tRNAs and the biogenesis of tRFs and tiRNAs in recent years (Fig. 1).

Fig. 1.

Classification and modification of tRFs. In the nucleus, the precursor tRNA is synthesized by RNA polymerase III and the 5’ end is cleaved by RNase Z or ELAC2 and tRF-1s are generated (Green). The mature tRNA transported into the cytoplasm is cleaved the D-loop or stem‒loop structure between the D-loop and the anticodon loop to generate tRF-5s by nucleic acid endonucleases (Yellow). Biogenesis of tRF-3s is from the cleavage of the T-loop (Red). Under stress induction, the anticodon loop of the mature tRNA is enzymatically cleaved to form 5’-tiRNAs and 3’-tiRNAs (Light blue)

Overall, the effect of tRNA modifications on translation is a current focus of attention, and the molecular mechanisms represented by methylation have been well elucidated [21]. Until the last few years, this delicate molecular mechanism of effect on the precise cleavage of tRNAs has received attention, and researchers have summarized the relationship with the biogenesis of tRFs [9, 22]. We summarize a few novel tRNA modifications identified in association with tRFs in this review. The enzymes involved in these modifications may be essential for future tRFs to perform important biological functions and have potential as biomarkers and therapeutic targets [23]. The result of tRNA modifications is often accompanied by production of rare bases, such as methylation with pseudouridylation, and induction of precise cleavage of RNAases [9, 24, 25]. Methylation is the most common type of tRNA modification and is often mediated by corresponding methylases [26, 27]. More common are methylation modifications of bases. For example, 5’-tRF-GlyGCC is elevated significantly in colorectal cancer, depending on upregulation of the demethylase AlkB homologue 3 (ALKBH3), which enables demethylation of m1A58, m3C32, or m3C47 [28, 29]. In addition, NOP2/Sun RNA methyltransferase 2 (NSun2)-mediated m5C48 and m5C49 were demonstrated to be associated with the biogenesis of certain tRFs [30]. Another nucleotide modification is 2’-O-methylguanosine (Gm). One study has shown that 5’-tRF derived from tRNA^Leu(CAA) with high Gm in the gut microbiota has anticancer activity [31]. Gm is widely distributed, including sites 18, 39, 44 and 54 [30]. Theoretically, methylation of tRNAs leads to their stabilization and is detrimental to the biogenesis of tRFs [32]. However, there are some exceptions, such as plants and microorganisms. For example, tRNAIle(GAU)-derived 3’-tiRNAs isolated from Ganoderma lucidum exhibit antitumour effects, which have been shown to be associated with tRNA modification of dihydrouracil (D) at site 19, m5U at site 52 and pseudouracil (ψ) at site 53 [33]. Other tRFs originate from hypomethylated tRNAs and are themselves hypomethylated [30]. For example, knockdown of tRNA methyltransferase 2 homologue A (TRMT2A) promotes m⁵U54 tRNA hypomethylation and induces angiogenin (ANG) overexpression, cleaving the tRNA anticodon loop region and promoting 5’-tiRNA biogenesis [34]. However, the above rules do not apply in specific biological contexts. Moreover, following mutation of the methylesterase dTrm7_34, which catalyses 2’-O-methylation of site 34, tRF-1 s and tRF-5s expression is reduced and tRF-3s expression increased [35]. In addition to methylation and demethylation, other specific modifications that affect the biogenesis of tRFs have been reported. In the presence of guanyltransferase, guanine is attached to the 5’ end of tRNA, which is associated with production of 5’-tiRNA [36]. Naturally, modified N6-isopentenyl adenosine (i6A) modification located at site 37 [36] significantly enhances the selectivity of tRNA-specific endonuclease deoxygenases and improves the efficiency of tRF biogenesis [37]. Liu et al. [38] found that in addition to catalysing binding of tRNA to leucine involved in translation, leucine-tRNA synthetase (LARS1) promotes acetylation of the 3’ end of tRNA to carry charge and induces the biogenesis of tRF-3s (Fig. 2).

Fig. 2.

tRNA modifications affecting the biogenesis of tRFs. Each nucleotide modification is labelled using the appropriate colour. The origin of a few modifications is known, and the modifying enzymes involved are drawn in different shapes next to the corresponding modified nucleotides. Hexagons represent methylesterases. Square represents demethylases. Triangle represents leucine-tRNA synthetase. Pentagram represents guanyltransferase

Enzymatic modification of tRNAs to regulate the biogenesis of tRFs is highly specific (Table 1). Compared to other ncRNAs, tRF biogenesis is still unclear and needs to be further investigated. tRNA modification is a direction of interest, and the enzymes involved have potential clinical applications and contribute to in-depth exploration of the biological functions of tRFs. We found numerous tRNA modifications concerning the biogenesis of tRFs, which have been systematically summarized [9], but their origin is unclear, and only a few modifying enzymes have been studied.

Table 1.

tRNA modifications associated with the biogenesis of tRFs

Modifications		Site	Modifying enzymes	tRNAs		Functions	Ref
				Name	Stability
Methylation	Gm	18	Unknown	tRNASer(GCT)	Up	Protect tRNA from ANG cleaving and inhibit the biogenesis of tRFs	[30]
				tRNASer(AGA)
				tRNASer(TGA)
Methylation	Gm	18	Unknown	tRNALeu(CAA)	Up		[31]
Methylation	Gm	34	dTrm7_34	tRNAGlu(CTC)	Down	Increase expression of tRF-5s	[35]
				tRNAGly(TCC)
				tRNACys(GCA)
Methylation	Gm	34	dTrm7_34	tRNAGln(CTG)	Down	Increase expression of tRF-1s	[35]
				tRNAPro(AGG)
Methylation	Gm	34	dTrm7_34	tRNAPro(CGG)	Up	Reduce expression of tRF-3s	[35]
				tRNAThr(AGT)
				tRNAGln(CTG)
				tRNAArg(TCG)
				tRNASer(GCT)
Methylation	Gm	39	Unknown	tRNATrp(CCA)	Down	Protect tRNA from ANG cleaving and inhibits the biogenesis of tRFs	[30]
Methylation	Gm	44	Unknown	tRNASer(AGA)	Down		[30]
				tRNASer(TGA)
				tRNASer(GCT)
Methylation	Gm	54	Unknown	tRNALys(TTT)	Down		[30]
				tRNALys(CTT)
Methylation	m5C	48	NSun2	tRNALys(CCT)	Up	Protect tRNA from ANG cleaving and inhibits the biogenesis of tRFs	[30]
				tRNAIle(AAT)
				tRNAThr(TGT)
				tRNAThr(CGT)
				tRNALys(CCT)
				tRNALeu(TAA)
				tRNATyr(GAT)
Methylation	m5C	49	NSun2	tRNASer(AGA)	Up	Protect tRNA from ANG cleaving and inhibits the biogenesis of tRFs	[30]
				tRNASer(TGA)
				tRNASer(GCT)
				tRNALys(TTT)
				tRNAPhe(GAA)
Methylation	m5U	52	Unknown	tRNAIle(GAU)	Down	Induce production of 3’-tiRNA with anticancer effects in Ganoderma lucidum	[33]
Methylation	m5U	54	TRMT2A	tRNAGly(GCC)	Up	Protect tRNA from ANG cleaving and inhibits the biogenesis of tRFs	[34]
				tRNAGlu(CTC)
Demethylation	m3C	32	ALKBH3	tRNAArg(CCT)	Down	Induce ANG cleaving of tRNA after demethylation and promote the biogenesis of tRF-5s	[28, 29]
				tRNASer(GCT)
Demethylation	m3C	47	ALKBH3	tRNASer(GCT)	Down		[28, 29]
Demethylation	m1A	58	ALKBH3	tRNAGly(GCC)	Down	Induce ANG cleavage and promote 5’-tRF-GlyGCC generation in colorectal cancer	[28, 29]
Other	i6A	37	Unknown	Unknown	Down	Enhance the selectivity of tRNA-specific endonuclease deoxygenases and improves the efficiency of tRF biogenesis	[37]
Other	Acetylation	CCA-OH	LARS1	tRNALeu(CAG)	Down	Charge the 3’ end of the tRNA and induce the biogenesis of tRF-3s.	[38]
Other	5′additions	5' end	Guanyltransferase	tRNAHis	Down	Induce anticodon cleavage in a tissue-specific manner and promote the generation of 5’-tiRNAs.	[36]

tRFs play a gene regulatory role in lung cancer

The regulation of gene expression plays an important role in the occurrence and development of lung cancer and is closely related to mRNA transcription [39]. miRNAs and circRNAs play pivotal roles during this process [40–42]. For example, our previous study suggested that miR-92a inhibits expression of the tumour-suppressor gene sprouty 4 (SPRY4), mediating the epithelial-mesenchymal transition (EMT) in non-small cell lung cancer [43]. CircVAPA, as a competitive endogenous RNA (ceRNA), sponges miR-377-3p and miR-494-3p to promote expression of insulin-like growth factor 1 receptor (IGF1R), which activates the protein kinase B (AKT) signalling pathway [44]. As miRNA-like small noncoding RNAs, tRFs are generally thought to silence downstream target genes at the posttranscriptional level [45]. A number of newly discovered tRFs have also been studied as miRNAs. We have previously summarized the role of tRFs in gene regulation and tumorigenesis, which is relatively rare in lung cancer [20]. Recently, several studies on the regulation of tRFs on downstream genes in lung cancer have been reported. However, recent studies have revealed a complex mechanisms of gene regulation by tRFs in lung cancer. By interacting with critical protein factors responsible for gene regulation, such as RNA-binding proteins (RBPs) and transcription factors (TFs), and altering their spatial structure and aggregation state, tRFs regulate transcription. We summarized the regulatory mechanisms of tRFs in lung cancer in recent years and present them here. (Table 2)

Table 2.

tRFs play a gene regulatory role in lung cancer

tRF name	Expression	Mechanism	Target		Function on downstream gene expression	Function on lung cancer	Type	Ref
ts-3676, ts-4521	Downregulated	piRNA-like function	PIWIL2		Affect downstream gene methylation and silences genes	Inhibition	LUAD	[73]
tsRNA-5001a	Upregulated	miRNA-like function	GADD45G		Form RISC complexes and silence downstream gene by binding to the 3’-UTR of target genes	Promotion	LUAD	[51]
AS-tDR-007872	Downregulated	miRNA-like function	BCL2L11			Inhibition	LUAD	[52]
5’-tRFCys	Upregulated	Interaction with RNA-binding proteins	nucleolin		Stabilizes transcripts by forming oligomers and promotes downstream gene expression	Promotion	Metastasis	[58]
AS-tDR-007333	Upregulated	Interaction with RNA-binding proteins	HSPB1		Interacts with promoter region histones for epigenetic modifications and promotes downstream gene expression	Promotion	LUAD	[61]
AS-tDR-007333	Upregulated	Interaction with transcription factors	ELK4		Promotes downstream gene expression through TFs binding to promoter regions	Promotion	LUAD	[61]
5’-IleAAT-8-1-L20	Upregulated	Unknown		Unknown	Unknown	Promotion	LUAD	[74]
tRF-Leu-CAG	Upregulated	Unknown		Unknown	Unknown	Promotion	LUAD	[75]

In the complex microenvironment of lung cancer cells, gene regulation of tRFs may lead to different phenotypes. We provide an exhaustive overview of the molecular regulatory mechanisms of tRFs in lung cancer that have been revealed thus far, including miRNA-like function and binding to RBPs and TFs. In addition, several studies have revealed other biological functions of tRFs in lung cancer, but their specific molecular interactions remain unclear. However, there is no doubt that they are also of potential research value.

miRNA-like function

In tumorigenesis, miRNAs form the RNA-induced silencing complex (RISC) with involvement of the Argonaute (AGO) protein to silence downstream target genes after directly binding to the 3ʹ-untranslated region (3’-UTR) of mRNAs. This mechanism has been systematically reviewed [46–48]. Our lab previously concluded that miR-199a regulates downstream target genes in lung cancer in such a manner [49].

tRFs derive from the precise cleavage of tRNAs, and due to the lack of intermolecular forces, tRFs do not have the neck-loop structure of tRNAs, which makes tRFs structurally similar to miRNAs. Therefore, tRFs have the ability to bind spatially to the 3’-UTR of mRNAs. In the development of lung cancer, miRNA-like functions of tRFs regulate critical transcription and play crucial gene regulatory functions. One study showed that tsRNA-5001a, which is upregulated in lung cancer, interacts directly with transcripts and that this interaction is highly similar to that of miRNAs, which is associated with an increased risk of recurrence and poor clinical prognosis after surgery and promotes proliferative capacity at the cellular level. The miMap database identified DNA damage 45G (GADD45G), which suppresses tumours through a signalling pathway [50], as a downstream target gene of tsRNA-5001a [51]. tsRNA-5001a might directly bind to the 3’-UTR of GADD45G, downregulate expression of GADD45G, and promote proliferation of lung adenocarcinoma (LUAD) [51]. Fan et al. [52] found that AS-tDR-007872, a tRF derived from 5’-tRNA, is downregulated in the tissues and plasma of lung cancer patients based on high-throughput sequencing and bioinformatics analysis, and overexpression of AS-tDR-007872 in NSCLC cell lines significantly inhibits cell proliferation, migration and invasion and promotes apoptosis. Based on TargetScan and KEGG websites, the open database Oncomine and real-time PCR, B-cell lymphoma 2-like 11 (BCL2L11), also called BIM, was identified as a downstream target gene of AS-tDR-007872. A previous study suggested that BCL2L11 lacks polymorphisms in non-small cell lung cancer associated with resistance to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) [53], and BCL2L11 is highly expressed in lung cancer. Thus, AS-tDR-007872 may act as a tumour suppressor by binding to the 3’-UTR of BCL2L11 (Fig. 3).

Fig. 3.

Downstream gene regulation of tRFs at the transcriptional level in lung carcinogenesis. (a) tRFs form a RISC complex with AGO and then bind to the 3’-UTR of the mature mRNA to perform gene-silencing functions. (b) In the nucleus, tRFs interact with intranuclear RBPs such as nucleolin to stabilize transcripts. In the cytoplasm, tRFs bind to extranuclear RBPs to enhance gene expression by regulating histone modifications. (c) tRFs enhance binding of transcription factors to gene promoter regions by regulating expression of transcription factors

The role of tRFs with miRNA-like functions in lung cancer depends on their downstream target genes. Due to structural similarities, miRNA-like action was the first transcriptional regulatory mechanism to be discovered and is often overlooked compared to miRNAs. Indeed, in-depth study of the miRNA-like roles of tRFs may help to find new ideas when miRNA-based biomarker studies or the development of miRNA-targeted drugs encounter bottlenecks. Although miRNA-like functions are the most studied mechanism of tRFs, they are still poorly understood in lung cancer. Hence, there is a need to further explore this mystery. In addition, distinguishing between regulation of the 3’-UTR by miRNAs and tRFs for the same mRNA is a challenge.

Interaction of tRFs with RNA-binding proteins

RBPs play crucial biological functions in lung cancer development by binding to RNA and mediating transcriptional regulation. For example, canonical RNA binding proteins CUGBP Elav-Like Family Member 6 (CELF6) and m6A reader YTH N6-methyladenosine RNA-binding protein 1 (YTHDF1) affect oncogenes in lung cancer [54, 55]. Some special RBPs, such as nucleolin, assemble ribonucleoprotein (RNP) with transcripts that activate transcription of oncogenes. A study suggested that this process may be regulated by ncRNAs, providing a new perspective on the transcriptional regulation of tRFs [56]. RBPs are involved in the gene regulation process of a few tRFs, which is not the case for miRNAs.

Metastatic lung colonization is a common route of lung cancer development intricately regulated by molecular mechanisms [57]. Liu et al. [58] found that after expression of 5’-tRF^Cys, a tRF derived from the 5-terminus of tRNAs carrying cysteine is associated with the degree of metastatic lung colonization in mice. This may be due to nucleolin binding to 5’-tRF^Cys in the nucleus. In the nucleus, 5’-tRF^Cys binds to nucleolin by two enrichment G motifs and promotes oligomerization to avoid cleavage of nucleolin-bound transcripts by exonucleases, providing a stable environment for transcript processing and translocation, such as Platelet-activating factor acetylhydrolase 1 beta 1 (Pafah1b1) and Methylenetetra-hydrofolate dehydrogenase 1 like (Mthfd1l), which are metabolic enzymes that play a key role in lung metastatic colonization. Overall, the 5’-tRF^Cys-nucleolin-Pafah1b1/Mthfd1l axis is a crucial mechanism that regulates metastatic lung colonization [58].

tRFs mediate a silencing mechanism similar to miRNAs in the cytoplasm but tend to influence the expression activity of transcripts through nuclear proteins in the nucleus. Due to the spatial effects of gene expression in eukaryotic cells, the effect of tRFs in the nucleus on gene expression is indirect, with possible mechanisms including providing a stable processing environment or involvement in the processing and transport of transcripts. Further exploration will facilitate the discovery of unknown biomarkers and therapeutic targets. However, we found that current studies on posttranscriptional regulation are not sufficient to support application of tRFs in clinical treatment of lung cancer (Fig. 3).

Normally, ncRNAs regulate downstream gene transcription through epigenetic modification via RBPs. Involvement of lncRNAs with RBPs in epigenetic modification, such as histone modification, has been previously reported and summarized [59]. Similarly, tRFs affect the development of lung cancer mainly through epigenetic regulation of transcription via extranuclear RBPs. Small heat shock protein-1 (HSPB1) has been shown to be an RBP that regulates gene transcription [60]. Recently, one study [61] found that AS-tDR-007333, a tRF with a length of 28 nt derived from cleavage of tRNA-Gly-GCC, promotes the proliferation and migration of NSCLC and is associated with poor prognosis in NSCLC by specifically binding to HSPB1. The UCSC Genome Browser and JASPAR database predict that HSPB1 might bind to histone H3-lysine 4-monomethylation (H3K4me1) and histone H3-lysine-27-acetylation (H3K27ac) sites in the promoter region of the downstream oncogenic factor Mediator complex subunit 29 (MED29), and the predicted results were verified by ChIP‒qPCR assays. In fact, previous studies have demonstrated that formation of H3K4me1 and H3K27ac affects transcription and carcinogenesis by regulating the activity of crucial oncogene promoters [62, 63]. Therefore, it is possible that AS-tDR-007333 regulates the promoter of MED29 and affects its expression by epigenetically modifying histones through the RNA-binding protein HSPB1 (Fig. 3).

Epigenetic modifications are typical of pretranscriptional regulatory processes. Previous studies on the regulation of epigenetic modifications of tRFs have focused on the effects of reverse transcription by transposons, on regulating the function of ncRNAs and on regulation of metabolism, such as lipid metabolism [45, 64]. In particular, epigenetic mechanisms of tRFs in lung cancer are poorly understood. A novel mechanism of pretranscriptional regulation of tRFs has been revealed that regulates histone methylation to impact DNA promoter binding and alter expression by RBPs, which provides new ideas for epigenetic modification of tRFs in lung cancer. There is no doubt that the regulation of histone methylation by extranuclear RBPs could be a potential research direction. Based on the refined biogenesis of tRFs and the comprehensive summary of their biological functions, we believe that tRFs also mediate other important epigenetic modifications in lung cancer, which will be one of the important directions for future research on the molecular mechanisms by which tRFs regulate gene expression.

Interaction of tRFs with transcription factors

TFs are a class of DNA sequence-specific binding proteins. Specific binding of TFs to target genes is achieved by nonspecific exploration of high-affinity DNA sequences. TFs bind to DNA by interacting with themselves and recruiting other cytokines [65]. Usually, TFs are associated with a complex regulatory network of signalling pathways in cancer, and they often form a circuit to ensure the proper functioning of the regulatory program [66]. By regulating expression of crucial signalling pathway molecules, TFs affect the occurrence and development of cancer, such as EMT [67]. A previous study mentioned that tRF^GluTTC inhibits adipogenesis by suppressing expression of lipogenic transcription factors [68]. Regarding tumorigenesis, TFs may act as downstream targets of tRFs. For example, runt-related transcription Factor 1 (RUNX1) inhibits the development of breast cancer by repressing downstream ts-112 expression in breast cancer [69]. However, the specific mechanism of the interaction between tRFs and TFs in lung cancer has not been extensively researched.

MED29 is also regulated by TFs. Another oncogenic mechanism of AS-tDR-007333 is activation of MED29 expression levels through ETS-like transcription factor 4 (ELK4), which is an important TF in cancer, activating transcription, regulating cell behaviour and promoting carcinogenesis [70]. JASPAR and UCSC database analyses suggested that the MED29 promoter region might interact with ELK4, which is highly expressed in lung cancer. The hypothesis was verified by ChIP-PCR and luciferase reporter assays [61]. This research demonstrated that tRFs exert a biological function by regulating expression of TFs in lung cancer, similar to breast cancer. However, the specific molecular mechanisms underlying the interaction between tRFs and TFs in lung cancer are unknown and need to be further explored (Fig. 3).

Notably, gene regulation mediated by TF binding to DNA promoter regions occurs prior to RNA transcription. The biological functions of TFs associated with ncRNAs in tumours have been extensively studied. For example, lncRNAs are epigenetically regulated through TFs by miRNAs [71], and miRNAs themselves are regulated by TFs in cancers [72]. However, little attention has been given to the function of tRFs in regulating lung cancer through TFs. AS-tDR-007333 influencing lung carcinogenesis via ELK4 regulation of MED29 reveals that tRFs interacting with TFs is a mechanism worthy of further investigation. This category of tRFs could be potential future biomarkers and novel therapeutic targets.

Other unrevealed regulatory mechanisms

In 2016, Yuri Pekarsky et al. [73] found that ts-3676 and ts-4521, which are downregulated in lung cancer samples, interacted with Piwi-like protein 2 (PIWIL2)-like piRNAs in an RNA immunoprecipitation (RIP) assay. This suggests that the molecular mechanisms of gene regulation by tRFs may be diverse (Table 2). However, these regulatory mechanisms have not been revealed in detail and need to be studied in more depth.

In a recent study, Sun et al. [74] found that 5’-IleAAT-8-1-L20, a 20 nt tRF-5 derived from tRNA^IleAAT-8−1, was upregulated in NSCLC and associated with proliferation. However, miRBase database analysis did not show a sequence similar to 5’-IleAAT-8-1-L20, suggesting that 5’-IleAAT-8-1-L20 may not have a similar miRNA structure. Luciferase reporter and RIP showed no significant negative correlation between expression of 5’-IleAAT-8-1-L20 and its downstream targets. Apparently, 5’-IleAAT-8-1-L20 regulates expression of downstream genes through a miRNA-independent mechanism [74]. Regrettably, the exact mechanism has not been elucidated. However, the conclusion that 5’-IleAAT-8-1-L20 is independent of miRNA-like effects points the way to future studies. Our group [75] previously identified that tRF-Leu-CAG is significantly upregulated in NSCLC tissues and cell lines and promotes the cell cycle and proliferative capacity of NSCLC. We found that tRF-Leu-CAG may act as a gene regulator by interacting with Aurora kinase A (AURKA), a common tumour-associated gene [76]. In addition, AURKA is highly expressed in LUAD [77] and associated with immunosuppression [77] and cell cycle abnormalities [78]. However, the expression level of tRF-Leu-CAG in NSCLC correlates positively with AURKA, demonstrating that tRF-Leu-CAG does not act as a silencing gene, as miRNA does. Our previous study suggested that miR-137 and miR-32 can target AURKA to inhibit NSCLC. Therefore, we hypothesized that tRF-Leu-CAG may interact with miRNA, but not mRNA, to affect NSCLC development [75]. The structure of tRFs is similar to that of miRNAs, but their interactions with miRNAs have received little attention. We propose a novel direction in which tRFs may affect lung cancer by regulating miRNAs. However, the manner and environment in which the two structurally similar small RNAs interact remains to be revealed.

Aberrant activation of signalling pathways is often present in most tumorigenesis [79, 80], including lung cancer [81]. In the progression of tumours such as gastric, colorectal and breast cancers, tRFs play important biological functions by targeting protein factors and regulating signalling pathways, such as the Wnt/β-catenin [82, 83], PTEN/PI3K/AKT [84], MAPK [85], Notch [86], hippo [87] and TGF-β1/Smad3 [88] signalling pathways. Unfortunately, the molecular mechanisms by which tRFs regulate signalling pathways in lung cancer have not received much attention, even though lung cancer is the most prevalent and fatal cancer worldwide. At present, only a few bioinformatics studies have revealed the signalling pathways involved in the regulation of tRFs in lung cancer. We provide a brief introduction below and summarize it in Table 3. There is no doubt that the effect of tRFs on signalling pathways in lung cancer is an unexplored direction with potential research value. We expect more researchers to focus on this important molecular mechanism.

Table 3.

Bioinformatics analysis of tRFs in lung cancer

Differential expression of tRFs	Expression profile analysis		Functional enrichment analysis		Ref
	Forecasting methods	Verification methods	Methods	Results
tRF-20-S998LO9D	MINTbase v2.0	qRT‒PCR	TargetScan, DAVID databases, GO and KEGG	Transcription-related pathways and Hippo signalling pathway, etc.	[91]
3P_tRNA-Arg-TCG-1-1, 5P_tRNA-Asn-GTT-2-3	TCGA-LUAD, GEO repositories, Sequencing	qRT‒PCR	GSEA, GO and KEGG	Proteoglycans in cancer, endocytosis and cell cycle, etc.	[92]
tiRNA-Lys-CTT-002 tRF-Ser-TGA-005 tRF-Val-CAC-010 tRF-Val-CAC-011	tRFdb database, Sequencing	qRT‒PCR	miRanda, TargetScan, GO and KEGG	Adrenergic signalling, cGMP-PKG signalling pathway, etc.	[93]
tRF-21-RKP4P9L0	Sequencing, tRFTar, TCGA-LUAD, GEPIA website	qRT‒PCR, Cell Counting Kit-8 and Transwell assays	GO, KEGG, Wikipathway, and Reactome Gene Sets enrichment analysis	Notch1	[94]

Bioinformatics analysis of tRFs in lung cancer

Gene regulation of tRFs in the complex microenvironment of lung cancer cells may lead to different phenotypes. We provided an exhaustive overview of the gene regulatory roles played by tRFs in lung cancer. Notably, the biological functions of tRFs in lung cancer have not been well revealed compared to those in other cancers. Bioinformatics combined with molecular biology techniques has become a critical tool for the discovery of tRFs that perform important functions [89]. Most of the expression profiling methods for tRFs in lung cancer involve database analysis combined with next-generation sequencing to initially identify differentially expressed tRFs. Subsequently, they are validated in cells or tissues using qRT‒PCR or northern blotting. We found a number of studies that only revealed the expression profiles and potential functions of tRFs in lung cancer with potential clinical applications, with further exploration needed. Therefore, we discuss this category of studies as well to provide a basis for further exploration of the mechanisms of tRF enrichment in lung cancer.

The MINTbase v2.0 database (http://cm.jefferson.edu/MINTbase/) is an effective way to reveal the expression profile of tRFs in cancer, as well as to predict tRFs as potential biomarkers, with primary data from The Cancer Genome Atlas (TCGA), specifically the ability to analyse and data tRFs in aggregate and reveal possible locations for parental tRNA modifications [90]. Ma et al. [91] predicted differentially expressed tRF-20-S998LO9D in lung cancer by the MINTbase v2.0 database and validated it by qRT‒PCR. However, since there is currently no uniform naming rule for tRFs, we cannot intuitively grasp the characteristics of tRFs based on their names in many studies. Moreover, tRFIDs have great differences in different databases, such as OncotRF (http://bioinformatics.zju.edu.cn/OncotRF/) and tRFdb (http://genome.bioch.virginia.edu/trfdb/), which are commonly used, leading to difficulties in data comparison. Gao et al. [92] combined tRFs into a unified expression form based on their origin and characteristics, which is a necessary method to facilitate future prediction and exploration of features. In addition, based on TCGA-LUAD and Gene Expression Omnibus (GEO) repositories, they screened 6 tRFs with significant differential expression that are potential diagnostic biomarkers for lung cancer. Subsequently, by sequencing tRFs in blood samples from patients and healthy individuals, only two tRFs (3P_tRNA-Arg-TCG-1-1 and 5P_tRNA-Asn-GTT-2-3) showed significant differences and were included in the six overlapping tRFs. Their expression profile was similarly validated in the human lung cancer cell line ABC-1 by qRT‒PCR. Prognostic analysis shows that 3P_tRNA-Arg-TCG-1-1 and 5P_tRNA-Asn-GTT-2-3 may play a critical role in lung cancer [92]. In another study, Zhang et al. [93] evaluated the results of tissue sequencing and identified 284 novel types of tRFs not included in the tRFdb database, 34 of which are differentially expressed in lung cancer. qRT‒PCR verified 10 tRFs and tiRNAs selected, and 4 were found to be differentially expressed between LUAD and adjacent tissues, named tiRNA-Lys-CTT-002, tRF-Ser-TGA-005, tRF-Val-CAC-010 and tRF-Val-CAC-011 [93] (Table 3). Based on the biogenesis of tRFs, in one of our previous studies, we performed next-generation small RNA (smRNA) sequencing using tRNAs differentially expressed in lung cancer [75]. Wang et al. [94] similarly identified tRNA expression profiles in lung cancer by small RNA high-throughput sequencing and then investigated differentially expressed tRFs. Specifically, they used the tRFTar website (http://www.rnanut.net/tRFTar) to find downstream regulated tRFs in lung cancer differentially expressed tRFs based on the principle of gene regulatory functions of tRF mRNAs and analysed the overlap of target mRNAs with those in TCGA-LUAD through the Gene Expression Profiling Interactive Analysis (GEPIA) website (http://gepia.cancer-pku.cn/). Finally, they established the tsRNA-mRNA regulatory network and identified three tRFs (tRF-16-L85J3KE, tRF-21-RK9P4P9L0, tRF-16-PSQP4PE) that are most closely related to mRNAs. Since only tRF-21-RKP4P9L0 was significantly associated with prognosis, they verified the expression and function of tRF-21-RKP4P9L0 by qRT‒PCR, Cell Counting Kit-8 and Transwell assays using A549 and H1299 cell lines [94].

Although previously unknown tRFs potentially affecting lung cancer have been revealed, their specific mechanisms in lung cancer remain unclear. In the above studies, functional analysis of tRFs almost always involved GO and KEGG function enrichment analyses, and tRFs were found to be closely associated with cancer-related pathways, such as aberrant transcription, aberrant metabolism, immune escape, metastasis and cell cycle checkpoint derangement [91–94]. We compiled the methods and findings of functional analysis of tRFs in lung cancer involved in the above studies thus far in Table 3. tRFs have quite complex and important mechanisms in lung cancer and are considered valuable biomarkers with potential clinical applications. However, the in-depth mechanisms involved in the development of lung cancer should be further revealed, which may provide more reliable treatment options for the future clinical treatment of lung cancer.

tRF as a biomarker for clinical lung cancer diagnosis and treatment

Clinical detection and treatment of cancer based on biomarker development have proven to be more effective than conventional approaches. With regard to lung cancer, the emergence of biomarkers has greatly improved the accuracy of clinical diagnosis and promoted the development of precision therapy [95, 96]. As biomarkers for the diagnosis and treatment of lung cancer, ncRNAs have promising clinical application prospects [97, 98]. Moreover, studies have shown that the differential expression and gene regulation of tRFs in lung cancer indicate their potential as biomarkers [99, 100] (Fig. 4). Due to the paucity of research results on tRFs in lung cancer, it is not possible to develop sound therapeutic regimens for this purpose. Therefore, most tRFs with the potential to enter the clinic as biomarkers are currently used for diagnosis of lung cancer. Further insight into the regulatory mechanisms of tRFs is necessary.

Fig. 4.

tRFs as biomarkers for clinical lung cancer diagnosis and treatment

tRF biomarkers in plasma

The majority of tRFs as biomarkers are found in plasma [101]. A study has shown that the molecular label TRY-RNA developed based on noncanonical small noncoding RNAs in peripheral blood, such as tRFs, is superior to traditional miRNA biomarkers and will be used as a new indicator for diagnosis of lung cancer in the future [102]. Thus, identification of differentially expressed tRFs in plasma is an effective method for screening clinical biomarkers. As mentioned above, based on the potential diagnostic value of tRFs in plasma and the tsRNA-mRNA regulatory network, the clinical diagnostic value of tRF-16-L85J3KE and tRF-16-PSQP4PE was found, and it was revealed that tRF-21-RK9P4P9L0 is a potential therapeutic target. You et al. [103] analysed differentially expressed tRFs and tiRNAs in LUAD patient plasma samples by high-throughput sequencing and qRT‒PCR, and tRF-1:29-Pro-AGG-1-M6 and tRF-55:76-Tyr-GTA-1-M2 were selected as candidate biomarkers. In addition, differentially expressed tRF-31-79MP9P9NH57SD in serum samples of NSCLC patients was screened by stem loop qRT‒PCR [104]. These molecules correlate significantly with clinicopathologic features such as tumour staging and lymph node metastasis and enriched in tumour-related signalling pathways [103, 104]. Compared to tissue samples, plasma is a plentiful and readily available source, facilitating measurement of biomolecule expression within it. Therefore, plasma-derived biomarkers are commonly used biochemical indicators for clinical diagnosis. In the diagnosis and targeted treatment of lung cancer, more than just one criterion should be considered. Clinical application of tRFs in plasma is reflected in their combination with other biomarkers and may be superior to common miRNAs. This method will be another powerful tool in the future clinical diagnosis of lung cancer (Table 4).

Table 4.

Possible tRFs as biomarkers for lung cancer

tRF name	Source	Type	Function	Clinical application	Ref
tRF-16-L85J3KE tRF-16-PSQP4PE	Plasma	LUAD	Unknown (Differential expression)	Diagnosis	[94]
tRF-21-RK9P4P9L0	Plasma	LUAD	Metastasis	Therapeutic target	[94]
tRF-1:29-Pro-AGG-1-M6 tRF-55:76-Tyr-GTA-1-M2	Plasma	LUAD	Result in poor prognosis	Diagnosis	[103]
tRF-31-79MP9P9NH57SD	Serum	NSCLC	Unknown (Differential expression)	Diagnosis	[104]
tRF-Leu-TAA-005, tRF-Asn-GTT-010, tRF-Ala-AGC-036, tRF-Lys-CTT-049, tRF-Trp-CCA-057	Exosome	NSCLC	Unknown (Differential expression)	Diagnosis	[109]
tRF-1:30-Lys-CTT-1-M2	Exosome	Lung metastases	Metastasis	Lung metastasis monitoring	[110]

tRF biomarkers in exosomes

Exosomes are a class of extracellular vesicles approximately 100 nm in diameter. The prototype is a multivesicular body within the endosome, which is secreted extracellularly to form exosomes as the endosome is translocated intracellularly and eventually fused to the cell membrane [105]. Exosomes contain a variety of functional biomolecules, such as circRNA and miRNA, which play a key role in lung carcinogenesis and are strongly associated with drug treatment and resistance [106, 107]. These contents are good biomarkers in clinical practice. In addition, excellent delivery and targeting capabilities, low immunogenicity and good metabolic excretion make exosomes potential drug carriers for the treatment of lung cancer [108]. In addition to plasma, differentially expressed tRFs in exosomes have the potential to serve as biomarkers with clinical applications. Zheng et al. [109] identified five differentially expressed tRFs in exosomes from lung cancer by high-throughput sequencing and qRT‒PCR, named tRF-Leu-TAA-005, tRF-Asn-GTT-010, tRF-Ala-AGC-036, tRF-Lys-CTT-049, and tRF-Trp-CCA-057. In lung metastatic cells, tRF-1:30-Lys-CTT-1-M2 is overexpressed in exosomes and highly correlated with lung cancer pathological features [110]. This study demonstrated that tRFs in exosomes can be used as biomarkers for the clinical monitoring of lung metastasis in the future. Compared to miRNAs and circRNAs, tRFs in exosomes are not well studied, especially because both exosomes and tRFs are highly heterogeneous and may be difficult to standardize across lung cancer cases. In addition, the difficulty of exosome purification and its instability limit its clinical application. However, there is no doubt that tRFs in exosomes may play an important regulatory role in the development of lung cancer, and their potential to become clinical biomarkers deserves further exploration.

In conclusion, tRFs have high clinical application value as new biomarkers and therapeutic targets. In recent years, researchers have sought to explore methods to screen tRFs as biomarkers in lung cancer and have achieved preliminary results. The combination of bioinformatics and biological detection techniques is an effective means to screen serum or exosome differentially expressed tRFs in lung cancer patients, and further studies should be conducted to identify diagnostic and therapeutic biomarkers in the future.

Conclusion and perspectives

With the development of modern biotechnologies such as high-throughput sequencing, the complex regulatory mechanisms of tRFs in cancer have been revealed, suggesting the need for more in-depth studies to better understand the patterns of cancer development. Compared to gastric and breast cancers, there are few studies on the regulation of tRFs in lung cancer, but they are novel and have clinical applications.

In this review, we first summarized the relationship between tRNA modification and the biogenesis and function of tRFs, which may be a potential mechanism for regulating lung carcinogenesis development. It is well known that tRFs have powerful gene regulatory functions, which have been summarized previously [15]. However, until now, the mechanism of tRF gene regulation and other biological functions in lung cancer has not been systematically summarized. In addition, there are some bioinformatic analyses on the differential expression and potential functions of tRFs in lung cancer. Previously, our laboratory suggested that tRFs act in cancer by regulating ribosomal function and have a role in influencing drug resistance [45]. We believe that the regulatory mechanisms of tRFs are not limited to the gene level but may even act as tRNA analogues competing with parental tRNAs for ribosomal binding sites to affect protein synthesis and amino acid metabolism. In conclusion, we summarize the bioinformatics study of tRFs in lung cancer to provide a theoretical basis for further in-depth study of other undisclosed regulatory mechanisms. Although the gene regulatory mechanisms of tRFs in lung cancer have not been well elucidated, these molecules have potential clinical applications. Finally, we summarize the clinical applications of tRFs as novel biomarkers and therapeutic targets to provide a better solution for the clinical diagnosis and treatment of lung cancer.

As important biomolecules, ncRNAs play a key role in gene regulation. Studies have shown that ncRNAs interact with each other to form a complex regulatory network. It is worth noting that circRNAs, lncRNAs and other ncRNAs often function by interacting with miRNAs, which may be central to the regulatory network of ncRNA interactions [111]. As mentioned above, the structure and some of the functions of tRFs are similar to those of miRNAs, and a study suggested that tRF-mRNA and miRNA‒mRNA regulatory networks may be related [112]. In addition, our previous study revealed the possibility of tRFs interacting with miRNAs [75]. Therefore, tRFs have the potential to interact with other ncRNAs, though there is a lack of sufficient evidence, and further study is needed. The potential interaction of tRFs with ncRNAs enriches the regulatory network of ncRNAs and provides an important reference to better reveal the biological functions of tRFs.

In conclusion, the biogenesis and biological functions of tRFs remain unclear. The lack of knowledge and systematic studies of tRFs leads to difficulties in further research, and this ultimately limits clinical application of tRFs. The importance of a better understanding of tRFs in lung cancer attack cannot be overstated, and the value of potential clinical applications such as targeted therapies and diagnostic biomarkers has driven them to become a hot research topic. Therefore, exploring novel lung cancer regulatory mechanisms of tRFs and developing more effective biomarker screening methods are goals that researchers should consider.

Abbreviations

ψ

pseudouracil

3’-UTR

3’-untranslated region

AGO

Argonaute

ALKBH3

AlkB homologue 3

ANG

Angiogenin

AKT

Protein kinase B

AURKA

Aurora kinase A

BCL2L11

B-cell lymphoma 2-like 11

CELF6

CUGBP Elav-Like Family Member 6

ceRNA

Competitive endogenous RNA

circRNAs

Circular RNAs

D

Dihydrouracil

EGFR-TKIs

Epidermal growth factor receptor tyrosine kinase inhib-itors

ELK4

ETS-like transcription factor 4

EMT

Epithelial-mesenchymal transition

GADD45G

DNA damage 45G

GEPIA

Gene Expression Profiling Interactive Analysis

GEO

Gene Expression Omnibus

Gm

2’-O-methylguanosine

GO

Gene Ontology

GSEA

Gene set enrichment analysis

H3K27ac

H3-lysine-27-acetylation

H3K4me1

H3-lysine 4-monomethylation

HSPB1

Small heat shock protein-1

i6A

N6-isopentenyl adenosine

IGF1R

Insulin-like growth factor 1 receptor

KEGG

Kyoto Encyclopedia of Genes and Genomes

LARS1

Leucine-tRNA synthetase

lncRNAs

Long noncoding RNAs

LUAD

Lung adenocarcinoma

LUSC

lung squamous cell carcinoma

MED29

Mediator complex subunit 29

miRNAs

MicroRNAs

Mthfd1l

Methylenetetra-hydrofolate dehydrogenase 1 like

ncRNAs

Noncoding RNAs

NSun2

NOP2/Sun RNA methyltransferase 2

NSCLC

Non-small cell lung cancer

Pafah1b1

Platelet-activating factor acetylhydrolase 1 beta 1

PIWIL2

Piwi-like protein 2

RBPs

RNA-binding proteins

RIP

RNA immunoprecipitation

RISC

RNA-induced silencing complex

RNP

Ribonucleoprotein

RUNX1

Runt-related transcription factor 1

SCLC

Small cell lung cancer

SPRY4

Sprouty 4

TCGA

The Cancer Genome Atlas

TEM

Transmission electron microscopy

TFs

Transcription factors

tiRNAs

tRNA semimolecules

tRFs

tRNA-derived fragments

TRMT2A

tRNA methyltransferase 2 homologue A

YTHDF1

YTH N6-methyladenosine RNA binding protein 1

Author contributions

Zhongliang Ma, Weiwei Wang, Fan Wu presented conceptualization. Qianqian Yang, Wei Pan, Wei Meng searched literatures. Fan Wu wrote the main manuscript text. Fan Wu, QianqianYang, Wei Pan, Wei Meng prepared figures and tables. Zhongliang Ma, Weiwei Wang supervised the review. All authors reviewed the manuscript.

Funding

This study was funded by the Shanghai Science and Technology Committee (No. 20S11901300) and Yunnan Provincial Science and Technology Department - Kunming Medical University Joint Fund Key Project (202201AY070001-137). Open access funding was provided by Shanghai University.

Data Availability

Not applicable.

Code Availability

Not applicable.

Declarations

Competing interests

The authors declare no competing interests.

Informed consent

Not applicable.

Ethics declarations

Not applicable.