A comprehensive expression profile of tRNA‐derived fragments in papillary thyroid cancer

Abstract Background The incidence of thyroid cancer has been on a rise. Papillary thyroid cancer (PTC) is the most common type of malignant thyroid tumor and accounts for approximately 85% of thyroid cancer cases. Although the genetic background of PTC has been studied extensively, relatively little is known about the role of small noncoding RNAs (sncRNAs) in PTC. tRNA‐derived fragments (tRFs) represent a newly discovered class of sncRNAs that exist in many species and play key roles in various biological processes. Methods In this study, we used high‐throughput next‐generation sequencing technology to analyze the expression of tRFs in samples from PTC tissues and normal tissues. We selected four tRFs to perform qPCR to determine the expression levels of these molecules and make bioinformatic predictions. Results We identified 53 unique tRFs and transfer RNA halves (tsRNAs). The 10 most upregulated tRFs and tsRNAs were tRF‐39‐I6D3887S1RMH5MI2, tRF‐21‐2E489B3RB, tRF‐18‐JMRPFQDY, tRF‐17‐202L2YF, tRF‐17‐VBY9PYJ, tRF‐18‐YRRHQFD2, tRF‐21‐WE884U1DD, tRF‐41‐EX2Z10I9BZBZOS4YB, tRF‐39‐HPDEXK7S1RNS9MI2, and tRF‐20‐1SS2P46I. The 10 most downregulated tRFs and tsRNAs were tRF‐31‐HQ9M739P8WQ0B, tRF‐43‐5YXENDBP1IUUK7VZV, tRF‐38‐RZYQHQ9M739P8WD8, tRF‐25‐9M739P8WQ0, tRF‐33‐V6Z3M8ZLSSXUD6, tRF‐27‐MY73H3RXPLM, tRF‐26‐DBNIB9I1KQ0, tRF‐38‐Z9HMI8W47W1R7HX, tRF‐40‐Z6V6Z3M8ZLSSXUOL, and tRF‐39‐YQHQ9M739P8WQ0EB. qPCR verification of cell lines and tissue samples yielded results consistent with the sequencing analysis. As tRF‐39 expression showed the maximum difference between PTC cells and normal cells, we chose this tRF to predict targets and perform functional tRF and tsRNA enrichment analysis. Conclusion In this study, we provided a comprehensive catalog of tRFs involved in PTC and assessed the abnormal expression of these fragments. Through qPCR verification, tRF‐39‐0VL8K87SIRMM12E2 was found to be the most significantly upregulated tRF. Further tRF and enrichment analysis revealed that tRF‐39 was mostly enriched in the “metabolic pathways.” These preliminary findings can be used as the basis for further research studies based on the functional role of tRFs in patients with PTC.


| INTRODUC TI ON
Thyroid cancer is one of the most common malignant tumors in the endocrine system 1 and has registered an annual growth rate of 2% in recent decades. 2 Papillary thyroid cancer (PTC) is the most common type of malignant thyroid tumor, with more than 10 known subtypes that account for 85-90% of all thyroid malignancies. 3,4 Fortunately, most PTCs are relatively inert and the 10-year survival rate for patients undergoing a glandectomy or total thyroidectomy plus I131 treatment is greater than 95%. 5 However, in approximately 10% of patients with PTC, recurrence or metastasis occurs and requires further intervention. Formulating treatments for PTC requires a thorough understanding of the molecular mechanisms underlying the pathogenesis and progression of the disease. Although the genetic background of PTC has been extensively studied, relatively little is known regarding the role of small noncoding RNAs (sncRNAs) other than microRNA (miRNA). 6 An earlier study led to the discovery of a new class of sncRNA derived from tRNA, known as tRNA-derived fragments (tRFs). 7 tRFs have functions similar to those of miRNA, with the ability to directly bind to mRNA, inhibit protein translation, and cleave partially complementary targets. 8  shown that tRFs exist in hematopoietic cells, lymphocytes, and the blood circulation system, suggesting that these fragments may play important role in the immune response. [11][12][13] Previous studies have shown that tRFs play regulatory roles in various human diseases such as cancers, pathological stress injuries, metabolic diseases, viral infections, and diseases of nervous system. [14][15][16] However, the understanding of how tRFs regulate cancer progression remains unknown. 17 Previous studies have indicated that tRFs and transfer RNA (tRNA) halves are involved in cancers of breast, prostate, colorectum, liver, and pancreatic, [18][19][20] although the potential role of tRFs and tRNA halves in thyroid cancer is yet to be elucidated. In this study, we sought to explore how the expression and function of tRFs regulate cancer cells in patients with PTC.

TA B L E 1 Patient and tumor characteristics
from the tumor) were collected for tRF sequencing and immediately stored in liquid nitrogen.

| Types of tRNA-derived small RNA (tsRNA)
Small RNAs derived from tRNA (tsRNAs) refer to the nuclease, such As a sncRNA, the length of mature tsRNA generated by tRNA shearing generally ranges from 16 to 35 nt ( Figure 3).

F I G U R E 1
Statistics of known tsRNA length. The length distribution of mature tsRNA generated by tRNA shearing is shown in Figure 2. The length of tsRNA is mainly concentrated in the range of 16-35 nts F I G U R E 2 Statistics of the number of known tRNA. The known tRNA ratio distribution is shown in Figure 3: the light blue part represents the proportion of GlyGCC, the orange part represents GluCTC, the gray part represents ValAAC, the yellow part represents GluTTC, the blue part represents VAlCAC, the green part represents GlyCCC, dark blue The color part represents HisGTG, and the brown part represents other F I G U R E 3 Statistics of the number of known tsRNA. The known tsRNA ratio distribution is shown in Figure 4: the light blue part represents the proportion of 5′-half, the orange part represents 5′-tRF, the gray part represents i'-tRF, and the yellow part represents 3′-tRF, the blue part represents 3′-half

| tRF sequencing
As tRFs are short fragments, the library was constructed using the following protocol: the total RNA or purified tsRNA fragments were extracted from samples with the 3′ end and the 5′ connector joined successively. Then, complementary DNA (cDNA) was synthesized using RT, followed by PCR expansion. The target fragment library was recovered via gel extraction and sequenced on F I G U R E 4 tsRNA heat map of different expression levels between samples. Heat map indicating the expression levels of various tRFs and tiRNAs.Brick red represents high expression of tsRNA in the sample, navy blue represents low expression. Each square represents a gene, and its color represents the amount of expression of the gene. The higher the amount of expression, the darker the color the computer. Illumina HiSeq 2500 was used to sequence the raw data reads. The joints at both ends of the reads were removed and low-quality reads and reads with a fragment length < 15 nt were also removed to complete the initial data filtering and to obtain high-quality data (clean reads). Clean reads were compared with the reference genome to obtain a genome-wide read distribution map, and were annotated according to sncRNA classification. The expressions of identified tsRNAs and cluster tsRNAs were calculated, and tsRNAs were analyzed for differential expression between samples.

| Quantitative RT-PCR (qRT-PCR)
We selected four tRFs for PCR verification. The sequences of the four tRFs are shown in Table 2

| tRF sequencing
We detected thousands of differentially expressed tRFs in PTC through high-throughput next-generation sequencing technology.

| Predicted targets of functional tRF and enrichment analysis
The target genes of tsRNAs that were produced from all three software programs and were used in the study can be seen in Figure 6. For target gene prediction, tRF-39-0VL8K87SIRMM12E2 was predicted to have 553 target genes by miRanda, RNAhybrid, and TargetScan. KEGG analysis was also performed on the target genes and an example of the KEGG channel analysis results is shown in Figure 7. tRF-39-0VL8K-87SIRMM12E2 was mostly enriched in the "metabolic pathways" section of the figure. Lastly, a candidate target gene KEGG pathway bubble chart showing the proportion of enriched differential genes in the background genes of the pathway is presented in Figure 8.

| D ISCUSS I ON
As sequencing technology has improved, research into sncRNAs has progressed and led to the discovery of tRFs. 24 Since the initial  In addition, RUNX1 may inhibit ts-112 to prevent hyperproliferation of breast cancer epithelial cells, thereby confirming its role in maintaining the breast epithelium. 30 In gastric cancer, -5034-GluTTC-2 is downregulated in gastric cancer tissue and plasma, and its level is significantly related to tumor size. The overall survival rate of patients with low -5034-GluTTC-2 expression is significantly lower than that of patients with higher expression. 31 These results indicate

| CON CLUS ION
In this study, we provided a comprehensive expression profile of tRFs in PTC and identified several potential avenues of further F I G U R E 8 Candidate target gene KEGG pathway bubble chart. In the figure, the abscissa indicates the ratio of enriched differential genes to the background genes of the pathway, and the ordinate indicates the pathway name; the size of the dots in the graph indicates the number of enriched differential genes, and the color indicates the p value research, including the biological role and marker potential of tRFs in thyroid cancer.

ACK N OWLED G M ENTS
This study was supported by the Social Development Foundation

CO N FLI C T O F I NTE R E S T
The authors indicated no potential conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are included in this published article [and its supplementary information files].