Plasma tsRNAs as novel diagnostic biomarkers for renal cell carcinoma

Abstract

Dear Editor, Early diagnosis greatly benefits renal cell carcinoma (RCC) patients with remarkable higher survival.1 Whereas, no appropriate liquid-biopsy biomarkers has been applied to RCC diagnosis in clinic so far.2 Transfer RNA-derived small RNAs (tsRNAs), a class of newly discovered noncoding RNAs, are stable and abundant in circulation, promising in noninvasive early diagnosis of cancer.3, 4 Here, we explore the diagnosis values and biological functions of plasma tsRNAs in RCC for the first time. The study was evaluated and authorised by the Ethics Committee of Nanjing Drum Tower Hospital (2021-582-01). The research route diagram was shown in Figure S1. First, to explore the distinct plasma tsRNA expression profile in RCC patients, we performed small RNA sequencing using RNA extracted individually from plasma of five RCC patients and five healthy controls. The tsRNA expression profile in plasma of RCC patients were obviously changed (Figure 1A), and among 657 tsRNAs detected, 71 tsRNAs were significantly altered (Figure 1B). Considering the tsRNAs expression levels (CPM > 10) and length (due to the length limitation, the length of tsRNAs should ≥ 17 nt when assessed by RT-qPCR), we selected 12 tsRNAs for further validation in an independent training set (30 healthy controls and 32 RCC patients) using RT-qPCR. The detection limits of RT-qPCR assay using tsRNA specific primers were evaluated by the standard curves developed with corresponding synthetic tsRNA oligonucleotides (Figure S2, Table S1). In this training set, 5 out of the 12 tsRNAs (tRF-19-DRMD5112, tRF-18-8R6Q46D2, tRF-17-884U1D2, tRF-17-8SOUPR2 and tRF-28-87R8WP9I1E0K) were significantly reduced in RCC patients, which was also consistent with the sequencing data (Figure 1C; Tables S2 and S3). Whereafter, the levels of these 5 tsRNAs were further confirmed in another larger validation set (99 healthy controls and 120 RCC patients), and showed a consistent trend (p < .0001 for all tsRNAs; Figure 2A and Table S4). Combining the training and the validation sets for analysis, the results were compatible (p < .0001 for all tsRNAs; Figure S3). Therefore, these data confirmed the stability of the expression pattern. Patient characteristics in the training set and validation set were summarised in Table S5. Plasma tRF-28-87R8WP9I1E0K level was decreased in patients with higher TNM grading (Table S6), suggesting its potential biological function in RCC progression. We then assessed the clinical usefulness of the five plasma tsRNAs for RCC. By ROC curve analysis, the AUC values for these tsRNAs ranged from 0.7487 to 0.8549 (p < .0001; Figure 2B). Among these tsRNAs, tRF-19-DRMD5112 showed the best performance (AUC: 0.8549, sensitivity: 80.92%, and specificity: 77.52%; Figure 2B). Furthermore, we assessed the values of these 5 plasma tsRNAs in early-stage RCC patients (stage I, 118 out of 152 RCC patients) diagnosis and figured out the range of AUC values from 0.7411 to 0.8650 (p < .0001; Figure S4). tRF-19-DRMD5112 also showed top performance (AUC: 0.8650, sensitivity: 80.51%, and specificity: 81.31%; Figure S4), which presented even higher diagnostic value in early-stage RCC patients compared to that in all RCC patients. To construct a tsRNA-based diagnostic model, we first performed LASSO-penalised logistic regression analysis using training and validation sets. Two tsRNAs, tRF-19-DRMD5112 and tRF-18-8R6Q46D2, were selected for the optimal model construction (Figure S5). However, in the following multivariate logistic regression analysis, tRF-19-DRMD5112 was the only tsRNA independently correlated with RCC (p < .001; Table S7). Taken together, the single expression level of tRF-19-DRMD5112 in plasma could be used as the diagnostic model for RCC. To further assess the diagnostic value of the model, a testing set (31 healthy controls and 43 RCC patients, Table S8) was utilised for external validation. Consistent with the results above, plasma tRF-19-DRMD5112 expression was significantly decreased in RCC patients (p < .001; Figure 2C) and could effectively distinguish RCC patients from healthy controls (AUC: 0.8882, sensitivity: 88.37%, and specificity: 77.42%, p < .001; Figure 2D). In addition to the diagnostic value, we further explored the potential function of these 5 tsRNAs in modulating RCC biological behaviours. Expression levels of tRF-19-DRMD5112 (17/21, p = .0028), tRF-18-8R6Q46D2 (13/21, p = .0281), and tRF-17-8SOUPR2 (17/21, p = .0032) expression levels were obviously higher in RCC tissues than that in paired normal tissues, while tRF-28-87R8WP9I1E0K (16/21, p = .0011) was obviously decreased in RCC tissues (Figure 3A). Characteristics of enrolled RCC patients were listed in Table S8. Furthermore, we constructed corresponding mimics of the five tsRNAs and transfected them to RCC cells, respectively (Figure S6). EdU assay, colony formation assay, and sphere formation assay found that only tRF-28-87R8WP9I1E0K substantially reduced the proliferation and self-new abilities of RCC cells (Figure 3B–D). Transwell assay and cell scratch test showed that tRF-28-87R8WP9I1E0K could markedly decrease the migration ability of RCC cells (Figures 3E and S7). Moreover, RCC cells over-expressed with tRF-19-DRMD5112, tRF-18-8R6Q46D2, tRF-17-8SOUPR2 or tRF-28-87R8WP9I1E0K showed reduced invasion ability (Figure 3F). Taken together, these results demonstrated that tRF-28-87R8WP9I1E0K played a powerful role in suppressing tumour progression of RCC. Structural prediction showed that tRF-28-87R8WP9I1E0K was located in position 1−28 of tRNA-Glu-TTC, which was of the tRF-5c type (Figure 4A). Considering that tsRNAs might exert the miRNA-like function, repressing the translation of target mRNAs via binding to AGO proteins and seed sequence-based canonical target recognition,5-7 we used targetscan_custom (https://www.targetscan.org/vert_40/seedmatch.html)8 and RNAhybrid (https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid/)9 to predict target genes of tRF-28-87R8WP9I1E0K (Table S9). Gene Ontology (GO) analysis of these predicted targets identified multiple enriched GO terms including negative regulation of gene expression, transcription factor activity, protein kinase activity and so forth (Figure 4B, Table S10). KEGG and WiKi pathway analysis showed that tRF-28-87R8WP9I1E0K might be involved in cGMP-PKG signalling pathway, amphetamine addiction, energy metabolism and other pathways (Figure 4C, Table S10). Together, these results reminded us that tRF-28-87R8WP9I1E0K might modulate the tumour biology of RCC through above mechanisms. In conclusion, we revealed plasma tsRNA expression patterns in RCC patients for the first time, and identified tRF-19-DRMD5112 with favourable diagnostic values for RCC (especially for early stage), providing novel potential biomarker for accurate RCC diagnosis. In addition, tRF-28-87R8WP9I1E0K is demonstrated to possess powerful anti-tumour function, which warrants further exploration to develop a novel tsRNA-based therapeutic strategy for RCC. MD, WD and HG conceived and designed the study. MD, WD and HG wrote the manuscript. MD and WD performed the quantification experiments. MD, WM and XG performed cell experiments. MD, WD and WM performed the bioinformatic analysis and statistical analysis. WZ, WC, XG, YL, CJ and GL contributed to sample collection. All authors read and approved the final manuscript. Not Applicable. The authors declare they have no conflicts of interest. This work was supported by grants from the National Natural Science Foundation of China (81902571 to M.D., 82173160 to W.D., and 81972388 to H.G.), Nanjing Medical Science and technology development Foundation (ZKX22024 to M.D.). The study was evaluated and authorised by the Ethics Committee of Nanjing Drum Tower Hospital (2021-582-01), and all participants voluntarily signed an informed consent. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts