Dynamic PAH-Related Changes: A Dataset of tRNA-Derived Small RNA Transcriptome Across Multiple Organs Authors: Yusi Chen, Yi Tang, Tengteng Zhu, Fang Li, Jun Luo, Jingyuan Chen, Yingjie Tan, Tianyu Wang, Jianqiang Peng, Gang Jiang, Jiang Li Published: 2024 ABSTRACT AbstractPulmonary arterial hypertension (PAH) is a severe cardiopulmonary disease characterized by elevated pulmonary artery pressure and cardiac dysfunction, often leading to heart failure and even death. Despite the complexity of its mechanisms, the molecular basis of PAH remains unclear. tRNA-derived small RNAs (tsRNAs) are noncoding RNAs that play regulatory roles in various diseases. We collected plasma samples from PAH patients and healthy controls for small RNA microarray, and we established a monocrotaline-induced PAH rat model to collect right ventricular and lung tissues for tsRNA sequencing. Our analysis revealed 2,716 unique tsRNAs in human plasma and 4,733 in rat tissues, with a 7.84% overlap. Additionally, 204 tsRNAs were highly conserved across plasma, lung tissue, and right ventricle samples. The reproducibility of the expression profiles was confirmed through Pearson correlation and principal component analysis. KEGG pathway enrichment analysis indicated that tsRNAs are involved in key pathways, such as the MAPK and cancer signalling pathways. These datasets offer valuable insights for researching the epigenetic mechanisms underlying PAH and potential therapeutic targets.Subject terms: Sequencing, Cardiovascular diseases Background & Summary Pulmonary arterial hypertension (PAH) is a severe cardiovascular disease characterized by elevated pulmonary artery pressure, which increases the right ventricular load and can lead to right heart failure1. The health of humans is seriously threatened by PAH; in fact, the median survival period for newly diagnosed patients is just 7 years, and recent studies have shown 1-year and 5-year survival rates of 86.3% and 61.2%, respectively2. PAH affects the lungs, heart, and multiple other organs, causing complex pathological changes3. Despite advances in understanding its pathophysiology and ongoing developments in treatment strategies, the currently available therapies for PAH remain suboptimal. PAH is influenced by genetic, epigenetic, and environmental factors, as well as by changes in cellular structure and function. Key regulatory genes, such as BMPR2 and CAV1, have been identified, but much remains unknown about the epigenetic regulatory mechanisms occurring at both the transcriptional and posttranscriptional levels4–6. tRNA-derived small RNAs (tsRNAs) are a recently discovered class of noncoding small RNAs that are produced by the cleavage of tRNAs by specific nucleases, such as Dicer and angiogenin7,8. tsRNAs are present in various tissues, and their expression profiles, abundance, and modifications are associated with factors such as sex, race, tissue type, and disease state9,10. tsRNAs are rich in RNA modifications, and they play crucial roles in the posttranscriptional regulation of protein translation, ribosome biogenesis, transgenerational epigenetic inheritance, stem cell biology, cell proliferation, cell apoptosis, and immune responses7,11,12. Changes in the expression patterns of tsRNA have been linked to the development of several illnesses, such as autoimmune diseases, metabolic disorders, cancer, and cardiovascular diseases13,14. However, research on the relationship between tsRNAs and PAH is still limited. In order to explore the role of tsRNAs in PAH, we obtained plasma samples from both PAH patients and controls, and conducted small RNA microarray analysis. Considering that PAH affects multiple organs, we also established a monocrotaline-induced PAH rat model and collected right ventricular and lung tissues for tsRNA sequencing. Detailed sample information is provided in Table 1. A total of 18 samples were analysed, including 6 published samples15, with 3 biological replicates per group. Our study successfully characterized the tsRNA expression profiles in the plasma of PAH patients and healthy controls, as well as in the right ventricular and lung tissues of model rats. These results provide significant resources and establish a basis for future investigation into the epigenetic processes that underlie PAH. Table 1Sequence reads information.SampleTotalReadTotalBaseBaseQ30BaseQ30 (%)N_right ventricle_11152332257616610051663398789.67N_right ventricle_2913913745695685040907534689.52N_right ventricle_31094875154743755049286835090.03M_right ventricle_1900891245044560040539432990.00M_right ventricle_2755755937787795034007756190.00M_right ventricle_31010478850523940045323532689.71N_lung_1591149329557465027298432592.36N_lung_2575619028780950026675102192.68N_lung_3761256538062825035453100493.14M_lung_1573944528697225026600592492.69M_lung_2578946928947345026733090692.35M_lung_3614668230733410028224474591.84Total Read: Raw sequencing reads after quality filtering. TotalBase: Number of bases after quality filtering. BaseQ30: Number of bases of Q score more than 30 after quality filtering. BaseQ30 (%): The proportion of bases (Q ⩾ 30) number after quality filtering. Sequence reads information. Total Read: Raw sequencing reads after quality filtering. TotalBase: Number of bases after quality filtering. BaseQ30: Number of bases of Q score more than 30 after quality filtering. BaseQ30 (%): The proportion of bases (Q ⩾ 30) number after quality filtering. Methods Sample collection and processing Our study received approval from the Ethics Committee of the Second Xiangya Hospital and followed the criteria established in the Helsinki Declaration for experimental procedures. This study included 3 patients who were diagnosed with idiopathic PAH according to the 2022 ESC/ERS guidelines16 as well as with 3 age- and sex-matched healthy controls. The PAH patients had a mean pulmonary artery pressure of 55, 59, and 59 mmHg, with ages of 28, 59 and 21 years (one female, two males). The control group had ages of 25, 48, and 26 years (one female, two males), with normal pulmonary artery pressures. After obtaining informed permission, blood samples were taken from all subjects during right cardiac catheterization. The samples were then centrifuged to isolate the plasma. Additionally, a PAH rat model was established in male Sprague-Dawley (SD) rats weighing around 200 g. After one week of adaptive feeding, 20 rats were randomly assigned to either the control group (n = 10) or the PAH model group (n = 10). To establish PAH, the model group received an intraperitoneal injection of monocrotaline (MCT, Sigma, batch number: WXBC4737V) at a dose of 60 mg/kg. The control group received an equal volume of sterile sodium citrate buffer via intraperitoneal injection. All rats were kept in plastic cages and environmental conditions were adjusted to an alternating light/dark cycle of 12 hours per day, with temperature at 22 ± 2 °C and humidity at 55 ± 5%. Four weeks post-modeling, rats were anesthetized with intraperitoneal sodium pentobarbital (40 mg/kg), and hemodynamic parameters were evaluated using a pressure transducer and physiological recorder via a right heart catheter. Pulmonary artery pressure (PAP) was significantly elevated in the MCT-treated group compared to the control group (mean PAP > 25 mmHg), confirming the development of PAH. Additionally, right ventricular hypertrophy (RVH) was assessed by calculating the right ventricle to left ventricle plus septum (RV/LV+S) weight ratio, with the PAH group showing a significantly higher ratio, indicating RVH. Lung histopathology further confirmed the PAH-like phenotype, with pulmonary vascular remodeling observed in MCT-treated rats, including medial thickening of the pulmonary arterioles. After euthanasia, lung tissues and right ventricles were collected from rats in each group, rinsed with heparinized saline, promptly snap-frozen in liquid nitrogen, and stored at −80 °C for subsequent analysis. RNA extraction and preprocessing For each group, three samples of plasma (500 µL) were processed directly for RNA extraction, while right ventricular tissue (20 mg), and lung tissue (20 mg) were homogenized using a tissue grinder. Each sample was homogenized in 1 ml of TRIzol (Invitrogen) to extract total RNA, adhering to the manufacturer’s protocol. Denaturing agarose gel electrophoresis was used to determine the integrity of total RNA samples. The concentration and purity of the extracted RNA were determined using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, USA). The concentration of the extracted RNA ranged from 20 to 200 ng/µL. RNA purity was evaluated by measuring the absorbance ratio at 260 nm to 280 nm, with samples exhibiting ratios between 1.8 and 2.0 being selected for sequencing analysis. For downstream analysis, approximately 1000 ng of total RNA was processed for each sample. Given the high level of RNA modifications in tsRNAs, RNA samples were processed with the rtStar tsRNA Pretreatment Kit (Cat# AS-FS-005, Arraystar) to remove modifications, such as acetylation and methylation, that could obstruct small RNA library construction. The pretreatment involved converting 3′-aminoacyl (charged) to 3′-OH and removing 3′-cP (2′,3′-cyclic phosphate) to enable 3′ adaptor ligation, phosphorylating 5′-OH to generate 5′-P for 5′ adaptor ligation, and demethylating m1A and m3C to enhance reverse transcription efficiency. Small RNA microarray analysis of Human Plasma Samples The total RNA from the plasma were denatured with dimethyl sulfoxide (DMSO), and the 3′-OH ends were enzymatically labelled with Cy3 fluorescent. The tsRNA’s probe includes 5′-cap segment dye5′-cap segment. This labelling step facilitates the detection and quantification of small RNAs during hybridization on the Arraystar Small RNA Expression Microarray (8X15K), which includes over 14,000 probes for various small RNA biotypes. The labelled RNA samples were then hybridized to the microarray chips at 42 °C for 16–20 hours. Library Construction and Sequencing Procedures of Rat Tissues We produced small RNA libraries from total RNA extracted from pre-treated right ventricle and lung tissue samples using NEBNext® Multiplex Small RNA Library Prep Set for llumina® (NEB, Catalog number: E7300L). RNA was ligated to 3′ and 5′ small RNA adapters, followed by cDNA synthesis using Illumina′s proprietary reverse transcription and amplification primers. Polymerase Chain Reaction (PCR) amplification generated fragments spanning 134 to 160 base pairs, aligning with tsRNAs ranging from 14 to 40 nucleotides in length. The PCR conditions for library amplification were as follows: initial denaturation at 94 °C for 30 seconds, followed by 12–15 cycles of 94 °C for 15 seconds (denaturation), 62 °C for 30 seconds (annealing), and 70 °C for 15 seconds (extension). A final extension was carried out at 70 °C for 5 minutes, and the libraries were then held at 4 °C. Amplified cDNA was loaded onto a 2% agarose gel and run at 120 V for approximately 45 minutes to separate the small RNA fragments based on size. The bands corresponding to RNA fragments between 14 and 40 nucleotides were excised from the gel. The finalized libraries were evaluated using an Agilent 2100 Bioanalyzer, pooled in equimolar ratios based on these evaluations, and then sequenced. Purified libraries were stored at −20 °C, with precautions taken to avoid repeated freeze-thaw cycles. The single-stranded DNA molecules were generated by denaturing the pooled library DNA fragments with 0.1 M NaOH. These fragments were then inserted onto the reagent cartridge at a concentration of 1.8 pM. The libraries were sequenced on an Illumina NextSeq 500 platform with 50 cycles in single-end mode. The sequencing depth targeted approximately 10 million reads per sample. Microarray data acquisition and analysis After washing and drying, the chips were scanned via a laser scanner (Agilent Scanner G2505C). The scanner settings were optimized on the basis of the chip characteristics and the excitation wavelengths of the fluorescent dyes to ensure optimal signal intensity and image quality. The captured files were processed using Agilent Feature Extraction software (version 11.0.1.1). The data processing steps included signal intensity measurement, background correction, normalization, probe summarization, and quality control. Raw signal intensities were measured from the CEL files, capturing the fluorescence intensities for each probe on the microarray chip. Background correction was performed using the Robust Multi-array Average (RMA) method, which adjusts for background noise and improves the accuracy of expression measurements. This method assumes that background noise is uniformly distributed and corrects the intensity values accordingly. Data normalization was performed using Quantile normalization to ensure consistency across different samples and platforms. Following normalization, the probe-level data were summarized to generate gene-level expression values using the median-polish method as part of the RMA algorithm. This step combines probe intensities for each gene into a single expression measure. Several quality control measures were applied, including inspecting boxplots of normalized intensity values to ensure consistency between arrays and calculating the percentage of present calls across arrays using the Affymetrix MAS 5.0 algorithm to assess the overall quality of the data. Sequencing data quality control and processing FastQC was implemented to assess the integrity of the sequencing. Each read consists of a small RNA sequence at the 5′ end, followed by the 3′ adapter sequence ligated during library preparation. The 5′ and 3′ adaptor bases were removed via Cutadapt, and lengths between 14nt and 40nt were retained. Trimmed reads, which were aligned to mature tRNA sequences, were subsequently aligned to a maximum of one mismatch. The trimmed sequencing reads were aligned to the Rattus norvegicus genome (version RGSC 6.0/rn6) using Bowtie. The reads were aligned in a manner that allows for a small number of mismatches to ensure accurate mapping of small RNA sequences. The expression profiles of tsRNAs were obtained by analyzing the number of mapped reads, and differential expression was evaluated using the R package edgeR, which relies on these read counts. For the identification of tRNA and tsRNAs, we used annotations from the GtRNAdb (Genomic tRNA Database, UCSC). Expression profiling of tsRNAs The abundance of tsRNAs was evaluated from sequencing counts and normalized as counts per million total aligned reads (CPM). The expression levels of each tsRNA were assessed by counting the number of mapped reads and using the allocated identification. tsRNAs with CPM values below twenty in each specimen were excluded from the study. The amount of tsRNA was determined by using the following formula: The variable i denotes the i-th read mapped to the tsRNA domain. The variable n represents the total amount of reads mapped to the tsRNA domain. The variable ci denotes the number of occurrences of the i-th read. The variable mi indicates the count of tsRNAs produced from the i-th read, which can exceed one when multiple mismatches are allowed. The CPM value was calculated as: The “Total aligned reads” refer specifically to reads that were aligned to tRNA and pre-tRNA sequences. Reads that were mapped to other RNA types, such as miRNAs, mRNAs, or other small RNAs, were excluded from this calculation Fig. 1. Illustration of sampled subjects, sample collection, RNA extraction, and analysis workflow. Functional enrichment analysis Recent research indicates that tsRNAs have regulation mechanisms that are comparable to those of microRNAs. Hence, we utilised the publically available miRanda and TargetScan algorithms to forecast the target genes of the tsRNAs. The TargetScan threshold was configured at 50, based on the context score percentile. For miRanda, the criteria involved a maximum binding free energy below −20. The genes that were identified by both miRanda and TargetScan were selected for the investigation of tsRNA targets using the Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathway. KEGG pathway enrichment analysis was conducted on the top 10 tsRNAs that were upregulated and downregulated, revealing significant conservation across the three tissues. Figure 2. depicts the top 10 pathways identified for each tissue. The upregulated tsRNAs were enriched in the MAPK signalling pathway, whereas the downregulated tsRNAs were enriched primarily in cancer signalling pathways and the MAPK signalling pathway. We investigated the associations between these pathways and the tsRNAs that were upregulated and downregulated. KEGG Pathway Enrichment Analysis of Differentially Expressed tsRNAs. This Fig. illustrates the KEGG pathway enrichment analysis for both up-regulated and down-regulated tsRNAs. The upper section represents the up-regulated tsRNAs, while the lower section represents the down-regulated tsRNAs. Data Records The raw data files associated with this study are available in the NCBI Gene Expression Omnibus (GEO) database under accession numbers GSE228750, GSE245790, and GSE270791. These datasets include both microarray and RNA sequencing data, covering tsRNA expression profiles from human plasma and rat lung and right ventricular tissues. The microarray expression profiles of human tsRNAs are provided under GSE228750 in CEL files (Affymetrix microarray format), which contain raw intensity values for each probe17. The RNA sequencing expression profiles of rat right ventricular tsRNAs are available under GSE245790 in FASTQ files containing raw sequencing reads18. Similarly, the RNA sequencing expression profiles of rat lung tissue tsRNAs are available under GSE270791 in FASTQ files containing raw sequencing reads19. These datasets are accompanied by metadata detailing the experimental conditions, sample preparation, and technical validation metrics, ensuring reproducibility and facilitating further analysis. Technical Validation Sequencing quality control The Illumina sequencing platform was used to create the FASTQ format data files. For the purpose of evaluating the quality of sequencing, a plot displaying the quality scores was generated for each individual sample. The quality score (Q) has a logarithmic relationship with the basecalling error probability (P) as follows: This relationship enables the assessment of the accuracy of each base call in the sequencing data. Reproducibility validation Correlation analysis was conducted on the 18 human and rat samples (Fig. 3A and B) to assess the repeatability of the biological replicates in the sample set. The majority of the biological replicates had strong correlation coefficients, according to the correlation heatmap (Fig. 3C and D). Furthermore, PCA revealed that the majority of biological samples grouped together (Fig. 3E and F). These results provide solid proof of the validity and consistency of the research data. Assessment of reproducibility across biological replicates. (A) Venn diagram illustrating conserved tsRNAs between human beings and rats. (B) Venn diagram illustrating conserved tsRNAs among plasma, right ventricle, and lung. (C,D) Heatmap of correlation of human beings and rats samples respectively. (E,F) PCA analysis of human beings and rats samples respectively. Supplementary information Data access note Acknowledgements This work was supported by the National Clinical Key Specialty Construction Project Cardiovascular Department (Cardiopulmonary Function Rehabilitation Direction) (Z2023123), Renshu Fund Project of Hunan Provincial People’s Hospital (RS2022A12) and Natural Science Foundation of Hunan Province (2022JJ30067). Author contributions Y.C. and J.L. designed the experiments. T.Y., T.Z., F.L., J.C. and Jun.L. collected the samples and constructed a sequencing library. Y.C. and J.L. analysed the data and wrote the paper. Y.T., J.G. J.P. and T.W. revised the paper. All the authors reviewed the manuscript. Code availability With the exception of those that are expressly included in the text and procedures, every software used in this investigation is freely accessible. This research did not use any customized scripts or procedures for the management and validation of datasets. The following software tools were employed within standard workflows. GtRNAdb provided the sequences of cytoplasmic tRNA14,20. The web tool Bioinformatics (www.bioinformatics.com) was used to construct PCA, correlation analysis, pie charts, Venn diagrams, and radar maps. Other software tools used include FastQC for sequencing quality assessment; Agilent Feature Extraction software for chip image analysis; These tools were run through standard pipelines and workflows commonly used in the field. No customized or proprietary scripts were developed for this study. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yi Tang, Email: da123zhiruoyu@163.com. Gang Jiang, Email: jianggang68@163.com. Jiang Li, Email: lijiangcs@csu.edu.cn. Supplementary information The online version contains supplementary material available at 10.1038/s41597-024-04115-0. References Associated Data Data Citations Supplementary Materials Data access note Data Availability Statement With the exception of those that are expressly included in the text and procedures, every software used in this investigation is freely accessible. This research did not use any customized scripts or procedures for the management and validation of datasets. The following software tools were employed within standard workflows. GtRNAdb provided the sequences of cytoplasmic tRNA14,20. The web tool Bioinformatics (www.bioinformatics.com) was used to construct PCA, correlation analysis, pie charts, Venn diagrams, and radar maps. Other software tools used include FastQC for sequencing quality assessment; Agilent Feature Extraction software for chip image analysis; These tools were run through standard pipelines and workflows commonly used in the field. No customized or proprietary scripts were developed for this study.