Serum tsRNA as a novel molecular diagnostic biomarker for lupus nephritis

Systemic lupus erythematosus (SLE) is an autoimmune illness with a lifelong toll on the body. Meanwhile, approximately 30% of SLE patients develop lupus nephritis (LN).1 However, clinical diagnosis of LN still relies on the "combination" of clinical symptoms, such as all proteins in the 24-hour urine of the patient (24-hour proteinuria) and anti-double stranded DNA antibodies (anti-dsDNA).2 tRNA-derived fragments (tsRNAs) are novel noncoding small RNAs (14∼40 nt in length) formed by the cleavage of mature tRNAs or their precursors under the action of enzymes or stress.3 tsRNAs have earlier biogenesis than most RNA and protein synthesis and are abundant in serum (> 50%) than miRNAs.4, 5 Hence, we researched and identified novel serum tsRNA molecular marker for the diagnosis of LN, with prospects to become noninvasive liquid biopsy tool for LN diagnosis. The sera of LN patients and healthy controls (HCs) were acquired at “Drum Tower Hospital affiliated with Nanjing University School of medicine” (Table 1). Thoroughly following the 1997 American Society of Rheumatology (ACR) , guidelines, signed consent forms from patients and approval from ethics committee (ID: 2020-327-01) were obtained. Serum RNA was extracted with TRIzol and treated with a rtStart™ tRF and tiRNA pretreatment kit by Arraystar Inc., USA, before sequencing. cDNA was constructed prior to quantitative reverse transcription polymerase chain reaction (RT–qPCR).6 When mapping the sequencing results, only one mismatch was allowed, and the differentially expressed tsRNAs (fold change > 10, P value < .01) in LN were selected for subsequent exploration (Figure 1A; and Table S1). tsRNAs of the tRF-1 type were more abundant in LN serum (Figure. 1B), which may be associated with systemic damage and preference of splice sites of angiopoietins in lupus patients.7 Venn analysis showed that 193 tsRNAs were present simultaneously in both groups (Figure 1C). The length range of tsRNA fragments was enriched at bases 20-23 nt and 30-32 nt (Figure 1D). Finally, we cluster analysed the 10 highly and 12 lowly expressed tsRNAs in LN (Figure 1A) and selected the highly expressed tsRNAs for marker screening (Figure 1E; and Tables S2, S3). RT–qPCR confirmed overexpression of tRF-Ala-AGC-2-M4 and tRF-Gly-TCC-1-M3 in the LN compared to HC (training set, Figure 1F) using agarose gel electrophoresis and sequencing of T vector clones to confirm the specificity of RT–qPCR (Figure S5). A standard curve of quantitative CT values was transformed for tsRNA concentration (Figure S1). tRF-His-GTG-1 has elevated levels in SLE patients but not in LN patients; hence, it can be used to distinguish LN patients from SLE patients.8 In this work, the total number of reads per million for tRF-Ala-AGC-2-M4 in the serum of LN patients was higher than that of HC patients (Table S2); RT–qPCR confirmed that tRF-Ala-AGC-2-M4 was exuberantly expressed in LN patients than in HC patients (P < .01), with area under the curve (AUC) = .7558 (validation set, Figure 2A). However, tRF-Gly-TCC-1-M3 expression was not significantly different between HC and LN (Figure S2). For LN clinical diagnosis, renal biopsy is the gold standard but is not suitable for dynamic monitoring.9 Anti-dsDNA and 24-hour proteinuria commonly used indicators for the diagnosis of LN in the clinic have low sensitivities and are time-dependent.2 Further clinical analysis showed that tRF-Ala-AGC-2-M4 had a higher sensitivity (81.08 and 91.67, respectively) and AUC (.7045 and .7500, respectively) in the diagnosis of LN with negative anti-dsDNA or 24-hour proteinuria (Figure 2B,C; and Table S4). Surprisingly, when anti-dsDNA or 24-hour proteinuria was combined with tRF-Ala-AGC-2-M4, the AUC of the two for LN detection reached.8502 and .9178, respectively (Figure 2D,E; and Table S5). Correlation analysis showed that tRF-Ala-AGC-2-M4 had a strong positive correlation with C-reactive protein (CRP) and the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K), with correlation coefficient values of .3299 and .3270, respectively (Figure 2F). However, tRF-Ala-AGC-2-M4 has little correlation with other indicators (Figure S3A). According to the SLEDAI-2K, tRF-Ala-AGC-2-M4 was highly expressed in the severely active lupus group and positively correlated with the degree of lupus activity (Figure 2G; and Table S6). However, other indicators were not significantly different among the four LN groups (Figure S3B). Thus, high levels of tRF-Ala-AGC-2-M4 in the serum correlate with the degree of activity in LN and can be an indicator for clinical evaluation for LN. Furthermore, on RT–qPCR validation of HC and four other autoimmune diseases, we verified the specificity of tRF-Ala-AGC-2-M4 in LN (Figure 2H; and Table S7). These results indicate that serum tRF-Ala-AGC-2-M4 has a certain clinical discrimination ability and can be used as a potential diagnostic marker for LN. Structural prediction showed that tRF-Ala-AGC-2-M4 originates after cleavage at the T-loop of tRNA-Ala-AGC-1, which is of the tRF-3a type (Figure 3A). Meanwhile, the coexpression network of 123 highly expressed target genes of tRF-Ala-AGC-2-M4 was mapped (Figure 3B). Gene Ontology (GO) analysis showed that biological, cellular and molecular functions were enriched in “Response to endoplasmic reticulum stress” (GO0034976), “Focal adhesion” (GO0005925) and “Cell adhesion molecule binding” (GO0050839), respectively (Figure 3C). The LN patient body is in a state of “tension,” while tsRNAs also have a stress-induced production pathway. Kyoto Encyclopedia of Genes and Genomes (KEGG) probing confirmed the pathway assignment in the “Huntington disease” (hsa05016) signalling pathway. More importantly, the wiki pathway was enriched in the “Nephrotic syndrome” (WP4758) signalling pathway. Meanwhile, these 10 highly expressed tsRNAs in LN serum were enriched in the above pathway (Figure S4). LN is a form of glomerulonephritis, and massive autoantibody involvement of the kidney induces LN, which is closely related to secondary nephrotic syndrome.10 Bioinformatics tools validated that tRF-Ala-AGC-2-M4 can be crucial for the development of LN. In summary, our study is the first to report the superior diagnostic ability of tRF-Ala-AGC-2-M4 and confirms that tsRNA markers are specifically highly expressed in LN serum and validated by sequencing and RT–qPCR of serum tsRNAs from patients with LN. Further clinical evaluation showed that tRF-Ala-AGC-2-M4 has good diagnostic value for LN and is closely related to the severity of LN. Finally, bioinformatics tools also predicted the involvement of tRF-Ala-AGC-2-M4 in the nephrotic syndrome signalling pathway. These results may provide new insights for accomplishing the goal of early diagnosis of LN and can be of value for making treatment decisions. The authors acknowledge all the participants of the study for providing consent and serum. This study has been supported by the National Natural Science Foundation of China (No. 61971216) and the Key Research and Development Project of Jiangsu Province (No. BE2019603, BE2020768). All authors affirm no conflict of interest. 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.