Cell culture: JURKAT (ATCC, TIB-152), TOM-1 (DSMZ, ACC 578) and 697 (DSMZ, ACC 42) cell lines were kindly provided by Dr. Felipe Prosper from the University of Navarra (Pamplona, Spain). JURKAT, TOM-1 and 697 cells were maintained in RPMI-160 medium (GIBCO, 11875119) supplemented with 10% fetal bovine serum (FBS) (Atlanta Biologicals, S11550). Cell lines were incubated at 37 °C under 5 % CO2 with medium changes every two to three days until confluent. Cell lines were screened for mycoplasma contamination and used within eight passages.
Isolation of CD19+ B-lymphocytes from blood: 12 mL of blood were obtained from eight (8) healthy donors without any exclusion criteria. Blood was diluted 1:2 with sterile PBS and layered over with 15 mL of Ficoll-Paque PLUS (Fisher Scientific, 11778538). Leukocytes were collected according to Ficoll-Paque manufacturer instructions. Two leukocyte pools of 4 individuals each were created. To isolate CD19+ B-lymphocytes, previously collected leukocytes were resuspended in 10 mL of sterile autoMACS® Running Buffer (Miltenyi Biotec, Cat. No. 130-091-221). After centrifugation at 700 g for 10 min, the pellet was resuspended in 80 μL of sterile autoMACS® Running Buffer. 20 μL of Dynabeads™ CD19 magnetic microspheres (Miltenyi Biotec) were added to the resuspended pellet, and the solution was incubated at 4°C for 10-15 min. Subsequently, CD19+ B-lymphocytes were recovered using LS columns (Miltenyi Biotec) mounted on the autoMACS™ Pro separator (Miltenyi Biotec), according to manufacturer’s instructions. After consecutive washings with sterile PBS, the recovered CD19+ B-lymphocytes were resuspended in RPMI at 1x106 cells/mL.
RNA isolation and real-time RT-PCR: Total RNA was isolated from CD19+ B-lymphocytes and B-ALL cell lines (697 and TOM-1) using both TRI Reagent® BD (500 μL/sample) (Sigma Aldrich, T3809) and TRI Reagent® BD-compatible columns according to the manufacturer's instructions (Direct-zol RNA MiniPrep Kit, Zymo Research, R2053). RNA quantification (ng/µl) and quality (A260/A280 and A260/A230 ratios) were measured in NanoDrop™ ND-2000 (ThermoFisher Scientific). Samples were stored at -80°C until use.
mRNA cDNA was synthesized from total RNA (400ng) using random hexamer technique (Thermo Fisher Scientific, N8080127) and SuperScript® IV Reverse Transcriptase (SSIV) enzyme (ThermoFisher Scientific, 18090010), according to the manufacturer's protocol in a two-step reaction. Quantitative real-time PCR was performed for SIRT1 mRNA using SYBR™ Green (Thermo Fisher Scientific, Cat. No. 4472908) and both specific 5’ (5’-CAAGGCCACGGATAGGTCC-3’) and 3’ primers (5’- ATTGTTCGAGGATCTGTGCCA-3’).
miRNA cDNA was synthesized from total RNA (400ng) using miScriptII RT kit (QIAGEN, 218160) following manufacturer’s instructions. Expression levels of miR-34a-5p were determined by quantitative real-time PCR (qPCR) using the miScript SYBR Green PCR Kit (QIAGEN, 218075), a universal 3’ primer (5’-GAATCGAGCACCAGTTACG-3’) and a specific 5' primers (5’-TGGCAGTGTCTTAGCTGGTTGT-3’).
All samples were amplified in the QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). The 2-(∆∆Ct) method was followed to calculate the relative abundance of miRNA and mRNA compared with endogenous control expression, u6 for miRNA and GAPDH for mRNA 28.
Total protein extraction and western blotting: Protein extraction was performed from CD19+ B-lymphocytes and B-ALL cell lines (697 and TOM-1) using RIPA lysis buffer (100µl/sample) (ThermoFisher Scientific, 89900). Protein quantification was carried out similarly as described before 29. Total protein (30 μg) was loaded in 8% polyacrylamide gels and separated by SDS-PAGE in reducing conditions. Gels were transferred onto PVDF membranes (Amersham Hybond P 0.45, Merk, GE10600023) and blocked 1 hour in 1% v/v PBS-Tween20 and 5% w/v BSA. Antibodies against human SIRT1 (1:1000, Abcam, ab7343) and human GAPDH (1:5000, Abcam, ab128915) were diluted in 1% v/v PBS-Tween20 and 5% w/v BSA, and they were incubated overnight at 4°C. SIRT1 and GAPDH were immunodetected with the appropriate peroxidase-conjugated secondary antibodies (anti-mouse NA931V; anti-rabbit NA9340V; 1:5000; GE Healthcare). ECL™ Prime Detection Kit (GE Healthcare, 12316992) and ImageQuant LAS 4000 Imager (ExonBiotec, GE Healthcare) were used for Western blotting detection. Densitometric analyses were performed with ImageJ software 30, and sample values were normalized with endogenous GAPDH levels.
B-ALL cell transfections: 697 (B-ALL) and TOM-1 (B-ALL) cells were seeded at 1x106 cells/mL per well in 6-well culture plates, and they were transfected with 100 nmol/L human miR-34a-5p mimic (ThermoFisher Scientific, 4464067), or miRNA scrambled control (SCR) (ThermoFisher Scientific, 4464058). Transfection reactions were performed using siPORT™ NeoFX™ (ThermoFisher Scientific, AM4510) as described before 31. After 48 h, cells were collected for subsequent RNA and protein analyses.
Cell proliferation assay: 697 (B-ALL) and TOM-1 (B-ALL) cells were seeded at 1x105 cells/mL in 96-well plates. Cells were incubated for 24h and 48h in OPTIM-MEM medium. B-ALL cells were transfected with 100 nmol/L human miR-34a-5p mimic or 100 nmol/L SCR, with or without the addition of 0.1 µM doxorubicin (DOX) (DOXPLAX®, Laboratorio LKM S.A., Argentina). Cell viability was assessed using the XTT Cell Proliferation II kit (ATCC, 30-1011K), according to the manufacturer's protocol. Cell viability was measured by colorimetry using a semiautomatic plate reader (BioTek® Synergy HT).
Transcriptome datasets selection of B-ALL patients: Gene Expression Omnibus (GEO) repository was used to identify potentially relevant studies with transcriptomic data from B-ALL patients 32. In this repository, the following keywords and expressions were used:
- For miRNA search: [(Acute Lymphoblastic Leukemia) OR (ALL)] AND [(miRNA) OR (microRNA)] AND [(transcriptomics) OR (miRNA-seq) OR (microarray)].
- For mRNA search: [(Acute Lymphoblastic Leukemia) OR (ALL)] AND [(ARNm) OR (messenger RNA)] AND [(transcriptomics) OR (RNA-seq) OR (microarray)].
From all potentially relevant datasets, the eligibility conditions included: (i) publicly available transcriptomic data (gene expression microarray, RNA-seq); (ii) detailed information on the sample and the protocol; (iii) patients with B-ALL among those studied. We selected the following datasets that complied with these criteria:
- GSE109868 33: The study included three plasma samples from healthy children, and nine paired samples of pediatric ALL individuals, which accounts for three newly diagnosed patient samples (D), three complete remission patient samples (CR), and three relapse patient samples (RE). miRNA expression profiles were detected by Taqman miRNA Array.
- GSE45839 34: From the forty-eight bone marrow samples collected, thirty-six represented B-ALL samples with different chromosomal alterations. All samples were obtained at diagnosis. miRNA expression profiles were detected by miRXplore_v4.0 TM Microarray. Associated clinical data included age, cell of origin (T or B), white blood cell count, response to prednisone, and relapse status.
- GSE28460 35: A total of ninety-eight paired samples were collected from pediatric B-ALL patients at two different time points: diagnosis (n=49) and relapse (n=49). Gene expression profiles were obtained by Affymetrix Human Genome U133 Plus 2.0 Array.
- GSE18497 36: A total of eighty-two paired samples were collected from pediatric B-ALL patients at two different time points: diagnosis (n=41) and relapse (n=41). Gene expression profiles were obtained by Affymetrix Human Genome U133 Plus 2.0 Array.
Additionally, cBioPortal from the Cancer Genome Atlas (TCGA) repository was used to download miRNA and mRNA transcriptomic data from B-ALL patients 37,38. In this repository, the "Pediatric Acute Lymphoid Leukemia - Phase II (TARGET, 2018)" database was selected, which contains miRNA (miRNA-seq) and mRNA (mRNA-seq) raw data from a total of 181 B-ALL patients. Clinical data associated included gender, age, BCR-ABL status, cell of tumor origin, site of relapse, days to last follow-up, and 5-year survival.
miRNA bioinformatic analysis: For the GSE109868 and GSE45839 datasets, we employed the GEO2R online tool from the GEO repository to identify differentially expressed miRNA 32. Using the GSE109868 dataset, we compared "Healthy children vs. B-ALL newly diagnosed patients”; and “B-ALL newly diagnosed patients vs. patients in complete remission, vs. patients in relapse". In the GSE45839 dataset, we utilized “relapse status” as the comparison condition. The miR34a-5p values obtained were downloaded in “.csv” format for statistical analysis.
To study differentially expressed miRNA from TCGA data, rows with zero counts (no expression) were removed, and the differential expression analysis was conducted using the R package DEseq2 (version 1.36.0) 39. miRNAs were considered differentially expressed if the Log2 Fold Change (Log2 FC) was greater than 1, and the adjusted p-value (Bonferroni correction) was below 0.05. The results were downloaded in ".csv" format for statistical analysis.
mRNA bioinformatic analysis: To identify differential mRNA expression levels in the GSE28460 and GSE18497 datasets, the GEO2R online tool from the GEO repository was used 32. The comparison condition was based on the disease relapse status. Furthermore, the differential expression analysis of mRNA data from TCGA was conducted using the R package DEseq2 (version 1.36.0) 39. In this case, the comparison condition was based on the 5-year survival status, categorizing individuals as "alive" or "deceased". In all genes expression analysis, genes with both a Log2 FC value greater than 0 and p-values less than 0.05 were filtered, and the list of filtered genes with associated expression values was downloaded in “.csv” format for further analysis.
Filtered out of potential direct targets of miR-34a-5p: A total of 899 potential target genes of miR34a-5p were downloaded from the miRDB database of microRNA-mRNA interactions 40. The "Bioinformatics & Evolutionary Genomics" web server tool (https:/ /bioinformatics.psb.ugent.be/webtools/Venn/) was employed to compare these mRNAs with the filtered gene list obtained from the GSE28460, GSE18497, and TCGA dataset analysis. The results of common mRNAs were illustrated as a Venn diagram suing the R package VennDiagram (version 1.7.3.) 41.
Risk scoring system analysis: A previously described risk scoring model 42 was utilized to perform a multivariable Cox regression analysis based on the gene expression levels obtained from the TCGA, GSE28460 and GSE18497 dataset analysis. B-ALL patients were categorized in different scoring groups according to the level of risk (high/low) using the Cutoff Finder software 43. Kaplan-Meier (KM) survival analysis were employed to calculate the overall survival (OS) between the dichotomized (high-risk/low-risk) patient groups 42.
Statistical analyses: Mean and standard error of the mean (SEM) were calculated using either GraphPad Prism 9 or R programming language. Non-parametric statistical calculations were used for transcriptomic data analysis, and the analysis of variance (ANOVA) was employed for multiple comparison analysis. Spearman's rank coefficient of correlation was utilized to calculate the association between continuous variables. The impact of miRNA and mRNA expression in the overall survival (OS) was evaluated using the Kaplan-Meier (KM) method 42 and the Log-rank test. We used the R package survminer 44 to plot the OS. Statistical significance was set at a p-value less than 0.05.