tRNA-derived fragments as novel potential biomarkers for relapsed/refractory multiple myeloma

Background tRNA-derived fragments have been reported to be key regulatory factors in human tumors. However, their roles in the progression of multiple myeloma remain unknown. Results This study employed RNA-sequencing to explore the expression profiles of tRFs/tiRNAs in new diagnosed MM and relapsed/refractory MM samples. The expression of selected tRFs/tiRNAs were further validated in clinical specimens and myeloma cell lines by qPCR. Bioinformatic analysis was performed to predict their roles in multiple myeloma progression.We identified 10 upregulated tRFs/tiRNAs and 16 downregulated tRFs/tiRNAs. GO enrichment and KEGG pathway analysis were performed to analyse the functions of 1 significantly up-regulated and 1 significantly down-regulated tRNA-derived fragments. tRFs/tiRNAs may be involved in MM progression and drug-resistance. Conclusion tRFs/tiRNAs were dysregulated and could be potential biomarkers for relapsed/refractory MM. Supplementary Information The online version contains supplementary material available at 10.1186/s12859-021-04167-8.

rapidly advances in high-throughput sequencing technology and bioinformatics analysis in recent years, a new class of sncRNAs derived from tRNAs are gaining increasing attention.
This new class of tRNA derived fragments can be broadly classified into tRNA related fragments (tRFs) and tRNA halves (tiRNAs). tRFs are generated from mature or precursor tRNA and tiRNAs are generated by specific cleavage in the anticodon loops of mature tRNA [6]. According to their mapped positions on the precursor or mature tRNA transcript, tRFs/tiRNAs are subdivided into five types: tRF-5, tRF-3, tRF-1, tRF-2 and tiRNA. tRFs/tiRNAs are involved in diverse molecular processes such as gene silencing, protein translation, cell stress, and cell differentiation [7,8]. Though the complex biological functions of tRFs/tiRNAs require further elucidation, they can be summarized as three categories: translation regulation, epigenetic regulation and RNA silencing. These three categories have also been a key focus in cancer research of tRFs/tiRNAs in recent years.
Increasing evidence shows that tRFs/tiRNAs contribute to cancer development and progression. For example, tDR-0009 and tDR-7336, which were significantly upregulated in hypoxia conditions, have been found to induce doxorubicin resistance in triplenegative breast cancer [9]. In ovarian cancer, tRF-03357 was reported to promote cell proliferation, migration, and invasion by downregulating HMBOX1 [10]. For hematological malignancies, limited data showed that tRFs/tiRNAs may have cancer-associated functions in leukemia and lymphoma [11,12]. And so far, there are no reports on the role of tRFs/tiRNAs in the mechanism of recurrence and drug resistance of MM to our knowledge.
In this study, we explored the expression profiles of tRFs/tiRNAs in new diagnosed MM(NDMM) and R/RMM samples by RNA-sequencing, with Quantitative Real-time PCR(qPCR)validation. Then, we analyzed their biological functions to uncover their roles in the mechanisms of relapse and drug resistance of MM. This study may provide potential biomarkers and therapeutic targets for R/RMM.

Clinical specimens
Bone marrow specimens were obtained from patients with MM for research according to a protocol approved by ethics committee of the third Xiangya hospital. All patients provided their written informed consent to participate in this study.20 new diagnosed MM (NDMM) and 22 R/RMM patients were enrolled, respectively. All NDMM patients met the criteria for symptomatic multiple myeloma according to diagnostic criteria defined by National Comprehensive Cancer Network (NCCN). The diagnosis of relapsed/refractory disease was based on clinical symptoms, biochemical parameters and marrow evaluation. All R/RMM patients were in first relapse.

Cell culture
MM cell lines U266 and RPMI-8226 were kindly provided by the basic laboratory of Central South University Xiangya School of Medicine. Drug-resistant cell lines U266/ BTZ, and RPMI-8226/BTZ were induced through stepwise increase of drug concentrations. Cells were cultured in RPMI1640 (HyClone, Logan, UT, USA) supplemented with 10% FBS (ExCell Biology,Shanghai, China) and 1% penicillin-streptomycin (HyClone). All cells were incubated at 37 °C in 5% carbon dioxide.

RNA Extraction and Quantitative RT-PCR
Firstly, anti-CD138 MicroBeads (Miltenyi, Germany) were used to enrich plasma cells from bone marrow samples by magnetic activated cell sorting (MACS). The purity of separated plasma cells was identified by flow cytometry and was all above 90%. Total RNA was then extracted from enriched plasma cells and myeloma cells using TRIzol (Invitrogen, USA) according to the instruction manual. Prepared RNA was stored at − 80 °C. RNA samples were qualified by agarose gel electrophoresis and quantified by NanoDrop ND-1000 (NanoDrop, USA). RNA integrity number (RIN) was evaluated by Agilent BioAnalyzer 2100 and was all above 7. RNA concentration and purity were also assessed (Additional file 1: Table 1). qRT-PCR was performed using ViiA 7 Real-time PCR System (Applied Biosystems) and 2 × PCR Master Mix. The expression levels of each tRFs/tiRNAs were calculated and normalized by U6 small nuclear RNA (snRNA). The primers used are listed in Table 1.

Library preparation and sequencing
Before library preparation, total RNA samples from 5 NDMM and 5 R/RMM patients were pretreated to remove RNA modifications that interfere with small RNA-seq library construction. Then, cDNA was synthesized and amplified using Illumina's proprietary RT primers and amplification primers. Subsequently, ~ 134 to 160 bp PCR amplified fragments were extracted and purified from the PAGE gel. Finally, the prepared libraries were quantified by Agilent BioAnalyzer 2100 and sequenced using Illumina NextSeq 500. Image analysis and base calling were performed by Solexa pipeline v1.8 (Off-Line Base Caller soft-ware, v1.8).
Sequencing quality were examined by FastQC and trimmed reads were aligned allowing for 1 mismatch only to the mature tRNA sequences. The abundance of tRFs/tiRNAs were evaluated using their sequencing counts and normalized as counts per million of total aligned reads (CPM). The differentially expressed tRFs/tiRNAs were calculated  based on the count value with R package edgeR. Pie plots, Venn plots, Hierarchical clustering, Scatter plots and Volcano plots were constructed using R or Perl.

Functional Analysis of tRFs/tiRNAs
Gene targets of tRFs/tiRNAs were predicted using TargetScan algorithms [13]. DAVID Bioinformatics Resources 6.8 was used for Gene Ontology (GO) and Kyoto Encyclopaedia of Genes and Genomes (KEGG) function enrichment analysis [14]. All data were graphed using Cytoscape 3.7.2 and GraphPad Prism 8.0.1.

Statistical analysis
SPSS 23.0 software was used for statistical analysis. qPCR value was presented as mean ± standard deviation. KS test and Shapio-Wilk test were used to determine whether the data were normally distributed. For data that obeyed normal distribution, statistical significance was assessed using two-tail unpaired Student's t test. Otherwise, non-parametric tests (Mann-Whitney test) were used for statistical analysis. P value < 0.05 was considered significant.

Catalogue of tRFs/tiRNAs expression in MM
After Illumina quality control and 5' , 3'-adaptor trimmed, reads with length < 14nt or length > 40nt were discarded. The bar diagrams show the sequence read length distribution ( Fig. 1a, b). We also calculated the frequency of subtype tRFs/tiRNAs against the length of the sequence. The stacked bar charts show the length distribution of subtype (Fig. 1c, d). Clinical characteristics of all patients are shown in Table 2

Target gene prediction with bioinformatics tool
TargetScan algorithms were used to explore the putative roles of tRF-60:77-Thr-TGT-1 and tRF-1:22-Lys-TTT-1-M3 in MM. Through this strategy, we predicted 238 conserved targets and 159 conserved targets, respectively. The network diagrams constructed by Cytoscape show the genes (Fig. 4a, b).

Discussion
MM is an incurable hematological malignancy. Even patients who are initially sensitive to treatment will eventually develop relapsed and refractory disease. So far, many genes or ncR-NAs related to relapse and drug resistance of MM have been identified. For example, J Xia demonstrated that NEK2 induces autophagy-mediated bortezomib resistance by stabilizing Beclin-1 in multiple myeloma [15]. MicroRNA-497 was found to inhibit myeloma growth and increase susceptibility to bortezomib by targeting Bcl-2 [16]. B-Z Zhu reported that LncRNA HOTAIR activated the expression of NF-κB and promoted the proliferation of myeloma cells [17]. However, it is still unclear whether tRFs/tiRNAs are involved in the relapse and drug resistance of MM. Therefore, for the first time, we conducted a preliminary evaluation of the role of tRFs/tiRNAs in R/RMM