Dysregulation of microRNA (miRNA) expression contributes to the pathogenesis of several clinical conditions. The aim of this study is to evaluate the associations between miRNAs and childhood acute lymphoblastic leukemia (ALL) to discover their role in the course of the disease. Forty-three children with ALL and 14 age-matched healthy controls were included in the study. MicroRNA microarray expression profiling was used for peripheral blood and bone marrow samples. Aberrant miRNA expressions associated with the diagnosis and outcome were prospectively evaluated. Confirmation analysis was performed by real time RT-PCR. miR-128, miR-146a, miR-155, miR-181a, and miR-195 were significantly dysregulated in ALL patients at day 0. Following a six-month treatment period, the change in miRNA levels was determined by real time RT-PCR and expression of miR-146a, miR-155, miR-181a, and miR-195 significantly decreased. To conclude, these miRNAs not only may be used as biomarkers in diagnosis of ALL and monitoring the disease but also provide new insights into the potential roles of them in leukemogenesis.
The prevalence of cancer is 11–15/100.000 among children and leukemia is the most common malignancy with an incidence of 30.2%, hence a major cause of mortality and morbidity [
A total of 45 children with newly diagnosed and untreated childhood ALL and 15 age-matched control subjects (normal peripheral blood (PB) and bone marrow (BM) smears) were enrolled in the study. Patients were consecutively included from the inpatient oncology and hematology departments of 3 hospitals. Previously diagnosed leukemia cases were excluded from the study as the miRNA expression levels might have been altered as a result of previous treatment. ALL diagnoses were confirmed via a BM aspirate showing at least 30% blast cells, in accordance with the FAB classification. All patients were diagnosed according to standard morphological, cytochemical, and immunophenotypic criteria. Patients were treated primarily with Berlin-Frankfurt-Munster- (BFM-) based national ALL protocol. This protocol was modified slightly with regard to methotrexate dosing and cranial irradiation. Following the BFM-ALL 1995 protocol, risk groups were categorized as standard (SRG), intermediate (IRG), and high (HRG) risk groups based on their age, leukocyte count, immunophenotyping, cytogenetic changes, early response to prednisone therapy, and BM remission. Patients received four drug induction regimens consisting of prednisone, asparaginase, vincristine, and daunorubicin. Complete remission (CR) was defined as a normocellular marrow with less than 5% blast cells. Patients’ characteristics, such as age, sex, white blood cell (WBC) count, FAB classification, and treatment response, are available in Table
Demographics, laboratory results, and response to the treatment in ALL patients.
Sex | Female |
21 (48.8) |
Male |
22 (51.2) | |
Mean age (year old) | 6.8 ± 4.5 | |
White blood cells | Mean (mm3) | 47.930 |
Range (mm3) | 487–474.000 | |
Subtype | T-lineage: 9 | |
B-lineage: 34 | ||
Blast BM | Mean (%) | 85 |
Range (%) | 52–100 | |
Blast PB | Mean (%) | 51 |
Range (%) | 2–100 | |
|
||
Characteristic | Status |
|
|
||
Risk group | Standard risk | 13 (30) |
Intermediate risk | 20 (47) | |
High risk | 10 (23) | |
Steroid response | Good | 36 |
Poor | 7 | |
Response at 33 day | Remission | 43 |
Not remission | 0 | |
Survival | Alive | 39 |
Dead | 4 | |
Relapse | Yes | 0 |
No | 39 | |
Dead | 4 |
ALL: acute lymphoblastic leukemia, BM: bone marrow, PB: peripheral blood.
The cases were evaluated prospectively. The enrollment period for each patient in the study was 18 months, plus 2 years of followup. MicroRNA microarray profiling was performed on all patients at day 0 and in control cases. The significant miRNAs were confirmed by real time RT-PCR. Six months after treatment, significantly dysregulated and validated miRNAs at day 0 were again analyzed by real time RT-PCR. The change in miRNA levels over time was determined by comparing the ratio at day 0 with the ratio of those measured at 6 months. During this period, 4 patients died. Among our study population, 3 cases (2 from patient group and 1 from control group) were also excluded following RNA isolation process, due to insufficient signalization during microarray work. The significance of miRNAs that were obtained from PB samples was compared with BM samples. Aberrant miRNA expressions associated with the diagnosis, differential diagnosis, and outcome of ALL were evaluated.
RNA isolation was performed on each BM and peripheral blood (PB) samples obtained from both patient and control groups. Total RNA was isolated by using Qiazol, which was then followed by miRNeasy Mini Kit (Qiagen, Valencia, USA) as per the manufacturer’s instructions. Genome wide microRNA microarray profiling was performed by using a microRNA biochip platform (Febit, Heidelberg, Germany). The platform consisted of 1136 microRNA probes (Sanger, miRBase 12.0). In short, 0.7 micrograms of microRNA was labelled using miRVANA labeling kit (Ambion, USA) and then dried via SpeedVac (Thermo, Germany). Dried samples were then treated with 18 microL of hybridization buffer (Febit Biomed GmbH, Heidelberg, Germany) and placed into the biochip platform overnight. Following hybridization and washing, signals were measured. The signal enhancement procedure was processed with Geniom Real Time Analyzer (GRTA) and detection pictures were evaluated by using Geniom Wizard software. The signal intensities for all miRNAs were extracted from the raw data for each array. After background correction, the median signal intensity of the seven replicate intensity values of each miRNA was obtained. Normalization was conducted by using the freely available R software (
The RNA quality control was determined by using the NanoDrop ND-1000 in which the ratios of 230/260 and 260/280 were 2. The RNA integrity number was determined by using Agilent 2100 Bioanalyzer (Agilent Technologies) and was ≥7.
The array data including raw data has been deposited in Gene Expression Omnibus (GEO) with the accession number GSE56489.
Significantly dysregulated miRNAs were validated by Quantitative PCR (LightCycler 480, Roche Applied Science, Mannheim, Germany) after comparing the BM microarray miRNA expressions of patients with the controls. RNA was reverse-transcribed to cDNA by using a cDNA synthesis kit (Exiqon). For miRNA quantification, the miRCURY LNA Universal RT microRNA PCR system (Exiqon) was used in combination with the predesigned primers (Exiqon). A master mix was designed for each primer set in accordance with the recommendations of the real time RT-PCR setup for “individual assays,” suggested in the kit. The reaction conditions consisted of polymerase activation/denaturation at 95°C for 10 min. For miRNA quantification, 40 amplification cycles at 95°C for 10 sec and 60°C for 1 min were performed; this was then followed by signal detection. The Delta-Delta-Ct algorithm was used to determine relative gene expression, and SNORD48 and U6 were used for housekeeping genes.
According to mean values, only those miRNAs whose “fold change” value demonstrated
This study was approved by the Local Ethics Committee, and written consent was taken from all parents of children that participated in the study.
The patient group consisted of 43 cases with ALL in which 34 cases (79%) were B-lineage, and 9 cases (21%) were T-lineage ALL. The mean age in the ALL group was
miRNAs from PB samples were not correlated with BM samples. Significant miRNAs in PB of patients when compared to control cases were totally different to miRNAs in BM. Therefore, all analysis and considerations in different categories were made by using BM samples. We propose using BM samples for the miRNA analysis in hematological malignancies because they reflect the leukemic process more efficiently, when compared to PB. With reference to the controls’ (
Significant miRNA profile compared to control cases in the microarray study and validation results in real time RT-PCR. According to mean values, only those miRNAs whose “fold change” value demonstrated ± 2-fold or more expression difference were included in the study. miRNAs whose “False Discovery Rate” (FDR) corrected
miRNA | Microarray | High/Low Expression ( |
RT-PCR | High/Low Expression ( |
|||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Control | ALL | Ratio |
|
FDR
|
Control | ALL |
|
| |||
|
1.84 | 22.12 | 12.50 | 3.90 |
0.03 | 23/0 | 0.000112 | 0.007942 | 6.15 | 0.590 | 21/0 |
|
46.55 | 488.77 | 10.00 | 6.55 |
0.01 | 21/7 | 0.002949 | 19.01167 | 12.65 | 0.291 | 28/3 |
|
72.46 | 467.47 | 6.25 | 7.09 |
|
36/3 | 0.0468 | 48.42431 | 10.02 | 0.251 | 33/2 |
|
6.66 | 23.84 | 3.57 | 2.69 |
0.03 | 14/0 | 0.003653 | 0.013143 | 1.85 | 0.163 | 7/0 |
|
326.65 | 1135.73 | 3.45 | 8.31 |
1.79 |
35/2 | 0.068189 | 24.37251 | 8.48 | 0.008 | 35/1 |
|
170.68 | 515.25 | 3.03 | 7.93 |
1.79 |
32/2 | 0.063027 | 31.66987 | 8.97 | 0.009 | 34/0 |
|
104.93 | 298.17 | 2.86 | 6.54 |
0.04 | 25/4 | 0.012341 | 0.766604 | 5.96 | 0.123 | 35/2 |
|
12.04 | 33.64 | 2.78 | 6.95 |
0.04 | 21/0 | 0.023803 | 0.001278 | −4.22 | 0.209 | 0/0 |
|
661.35 | 1810.68 | 2.70 | 3.66 |
0.03 | 29/7 | 0.770556 | 80.63069 | 6.71 | 0.002 | 34/2 |
|
720.72 | 1713.49 | 2.38 | 2.67 |
2.88 |
30/0 | 0.245556 | 11.3962 | 5.54 | 0.009 | 33/4 |
|
34.21 | 81.96 | 2.38 | 1.05 |
0.01 | 25/2 | 0.000686 | 0.000024 | −4.84 | 0.199 | 0/0 |
|
962.76 | 2251.43 | 2.33 | 9.67 |
0.01 | 28/4 | 0.012433 | 1.250504 | 6.65 | <0.001 | 32/3 |
|
33.83 | 69.93 | 2.08 | 5.67 |
0.04 | 17/4 | 0.000523 | 0.000146 | −1.84 | 0.326 | 1/0 |
hsa-miR-640 | 91.53 | 45.07 | −2.03 | 7.18 |
0.04 | 3/30 | 0.050227 | 0.000498 | −6.66 | 0.191 | 0/0 |
hsa-miR-145 | 498.27 | 197.73 | −2.52 | 8.09 |
1.79 |
2/35 | 1.201444 | 14.55961 | 3.60 | 0.440 | 26/10 |
FDR: false discovery rate, ALL: acute lymphoblastic leukemia.
*“High/Low Expression (
Significant miRNAs after validation by real time RT-PCR at day 0 were given. Those miRNAs were reevaluated after 6 months of treatment and the expression change of all miRNAs except miR-128 was significant.
miRNA | RR | |
---|---|---|
Validated by RT-PCR (at diagnosis) | hsa-miR-128 | 11.396200 |
hsa-miR-146a | 24.372519 | |
hsa-miR-155 | 31.669875 | |
hsa-miR-181a | 80.630694 | |
hsa-miR-195 | 1.250504 | |
|
||
Real time RT-PCR (after 6 months) | hsa-miR-128 | 6.073660 |
hsa-miR-146a | 0.334798 | |
hsa-miR-155 | 1.201844 | |
hsa-miR-181a | 4.987600 | |
hsa-miR-195 | 0.206393 |
RR: relative ratio.
Heat map and cluster analysis in ALL patients and control cases (FDR
Diagram of significantly changed miRNAs in ALL patients after 6 months of treatment. The figure shows the change in expression levels of 4 significantly dysregulated miRNAs after 6 months of treatment. The change in miRNA levels over time was determined by comparing the ratio at day 0 with the ratio of those measured at 6 months by using quantitative real time RT-PCR.
The most significantly upregulated miRNAs and their fold changes in T-lineage ALL and B-lineage ALL are summarized in supplemental Table 1 (see Supplementary Table 1 in Supplementary Material available online at
In this study a total of 1136 miRNAs were studied in children with ALL using microarray platform to reveal their contribution with regard to diagnosis, classification, and treatment period. Microarray study revealed 15 miRNAs dysregulated, when compared to the control cases. The results of real time RT-PCR for 5 miRNAs (miR-128, miR-146a, miR-155, miR-181a, and miR-195) were consistent with our microarray results; however, the other 10 miRNAs (miR-548i, miR-708, miR-181b, miR-369-3p, miR-449a, miR-3121, miR-181a, miR-1323, miR-587, and miR-181a-2) contradicted the results of our microarray study. This discrepancy may be due to the small sample size (
Studies have shown that
Some miRNAs are already known to be related to specific subtypes of pediatric ALL. As reported by Schotte et al., genetic subtypes such as MLL-rearranged, TEL-AML1 positive, hyperdiploid, and drug-resistant leukemic cells display characteristic miRNA signatures in pediatric ALL [
The limitation of our study lies in the studying of expression levels in specimens containing both normal and blast cells where the results may be affected by the contribution of normal cells. The mean blast cell level in our study, however, was found to be 85% (range: 52–100) meaning this issue may be avoided.
When the age group and its characteristics of childhood leukemia are considered, our data could add important contributions to the literature. The first is the studying of the largest miRNA profile in ALL and the presentation of novel miRNAs associated with leukemogenesis; the second contribution is identifying miRNAs as discriminative of T-lineage versus B-lineage ALL. Moreover, our results confirmed the importance of certain miRNAs such as miR-128, miR-146a, and miR-181a in childhood ALL. The final and possibly the most important contribution is the prospective design of our study that we were able to evaluate miRNAs throughout a treatment period.
In conclusion, the discovery of miRNAs and their association with disease have provided valuable information on potential diagnostic and/or prognostic biomarkers, as well as monitoring the disease progression. In our study, miR-128, miR-146a, miR-155, miR-181a, and miR-195 were found to be significantly dysregulated which may help provide new insights into the diagnosis and prognosis of childhood ALL. Further studies, with larger subject numbers, are needed to clearly demonstrate the effect of miRNAs in leukemogenesis and its practical implications.
The authors declare that they have no conflict of interests regarding the publication of this paper.
Muhterem Duyu and Burak Durmaz contributed equally to this work.
The authors would like to thank the Scientific and Technological Research Council of Turkey (TUBITAK) for financial support under Grant 109S076.