Modulation of microRNAs expression in hematopoietic stem cells treated with sodium butyrate in inducing fetal hemoglobin expression

Abstract

Context Inherited hemoglobin diseases are the most common single-gene disorders. Induction of fetal hemoglobin in beta hemoglobin disorders compensate for abnormal chain and ameliorate the clinical complications. Sodium butyrate is used conventionally for fetal hemoglobin induction; it can be replaced by safer therapeutic tools like microRNAs, small non-coding RNAs that control number of epigenetic mechanisms. Objective In this study, we compared the changes in the microRNAs of differentiated erythroid cells between control and sodium butyrate treated groups. The objective is to find significant association between these changes and gamma chain up regulation. Materials and methods First, CD133⁺ hematopoietic stem cells were isolated from cord blood by magnetic cell sorting (MACS) technique. After proliferation, the cells were differentiated to erythroid lineage in culture medium by EPO, SCF, and IL3. Meanwhile, the test group was treated with sodium butyrate. Then, gamma chain upregulation was verified by qPCR technique. Finally, microRNA profiling was performed through microarray assay and some of them confirmed by qPCR. Result Results demonstrated that gamma chain was 5.9-fold upregulated in the treated group. Significant changes were observed at 76 microRNAs, in which 20 were up-regulated and 56 were down-regulated. Discussion Five of these microRNAs including U101, hsa-miR-4726-5p, hsa-miR7109 5p, hsa-miR3663, and hsa-miR940 had significant changes in expression and volume. Conclusion In conclusion, it can be assumed that sodium butyrate can up-regulate gamma chain gene, and change miRNAs expression. These results can be profitable in future studies to find therapeutic goal suitable for such disorders.

Keywords:

• Fetal hemoglobin
• hematopoietic stem cell
• microarray
• microRNA
• sickle cell anemia
• sodium butyrate
• thalassemia

Introduction

Synthesis of an abnormal βs-globin chain and the reduced rate of synthesis of normal β-globin chains result in Sickle cell diseases and β-thalassaemias, respectively (Galanello and Origa 2010, Pace et al. 2015, Rees et al. 2010). Sickle cell disease is a widespread group of hemoglobin disorders caused by substitution of single amino acid. Homozygous sickle cell anemia (HbSS) is the most common severe syndrome resulting from insoluble Hbs and formation of crystals when exposed to low oxygen tension (Madigan and Malik 2006). Due to the fetal hemoglobin (HbF) exclusion from interaction with HbS, and subsequently its exclusion from the sickle polymer formation, HbF is consider as the major genetic modulator of the hematological and clinical features of sickle cell anemia (Noguchi et al. 1988).

β-Thalassemia syndromes are a heterogeneous group of monogenic disorders worldwide, characterized by molecular mutations of the β-globin chain, which cause deficiency of β-globin chains and an excess of unmatched α-globin chains (Galanello and Origa 2010). Excess unbalanced α globin chains precipitate in erythroid precursors causing the severe ineffective erythropoiesis and intramedullary hemolysis (Mathias et al. 2000). The greater the α-chain excess, the more severe the anemia (Rund and Rachmilewitz 2005). Production of γ-globin, to a critical level helps to decrease the excess α chains and to ameliorate the condition in patients with β-chain abnormalities and obviate the regular transfusions requirement (Perrine 2008).

Fetal hemoglobin (HbF; α2γ2) is the major hemoglobin in the fetus, but it subsequently declines and adult hemoglobin, HbA (α2β2), becomes predominant (Singer et al. 2005). The main switch to adult hemoglobin occurs 3–6 months after birth; hence, clinical symptoms of the patients with sickle cell anemia and β-thalassemia major would become apparent at nearly that time (Stamatoyannopoulos 2005). In last few decades, hemoglobin switching has been the target of various investigations, and it is expected to lead to the development of useful therapeutic approach to thalassemias and hemoglobinopathies (Perrine 2008). Inducing γ-globin expression by even small increments can compensate reduced or abnormal β-chain expression and therefore inhibits the polymerization of Hbs in sickle cell anemia and ameliorates the clinical severity of the β-thalassemias (Wood et al. 1976).

Principally, we know today, some mechanisms that regulate globin gene activity during development (Stamatoyannopoulos 2005). A number of pharmacological agents are available as fetal hemoglobin inducers including hydroxyurea, 5-azacytidine, araC, butyrate, and other short chain fatty acids (Chou et al. 2015, Perrine 2011, Steinberg and Rodgers 2001). Although hydroxyurea is the only FDA approved drug for SCD treatment, but it is highly pleiotropic and does not solely modulate γ-globin gene expression (Costa et al. 2013). An oral butyrate derivative, sodium 2,2-dimethylbutyrate (SDMB) is one of these drugs, which found to stimulate γ-globin production in thalassemia and sickle cell anemia patients (Faller and Perrine 1995). There is a growing interest in butyrate derivations because its impact on epigenetic mechanisms, which leads to more specific and efficacious therapeutic strategies for the prevention and treatment of different diseases (Faller and Perrine 1995). Although the exact mechanism by which sodium butyrate affects gene transcription remained unknown, the common opinion is that this short-chain fatty acid induces γ-globin gene expression through the inhibition of histone deacetylases (Cao et al. 2004, Panja and Basu 2015). Currently, the main focus is on the development of more safe agents, which can increase HbF to prevent all complications of the diseases (Fathallah and Atweh 2006, Sankaran 2011).

MicroRNAs (miRNAs or miRs) are small (about 22 nucleotides in length) non-coding single-stranded RNA molecules (Minayi et al. 2014), which play key role in the regulation of gene expression (Zare et al. 2014). This interaction prevents protein production by suppressing protein synthesis and/or by initiating mRNA degradation (Valencia-Sanchez et al. 2006). Recently, miRNAs have been identified to regulate HbF expression in erythropoietic cells (Bianchi et al. 2009, Gabbianelli et al. 2010, Sankaran et al. 2011). Nevertheless, whether γ-globin mRNAs could also be directly targeted by certain miRNAs has not been reported yet (Azzouzi et al. 2011).

In the current study, we studied sodium butyrate-mediated induction of HbF by analyzing epigenetic and molecular profiles of erythroid cells derived from CD133 + HSC. We selected sodium butyrate since it can consistently affect the gamma chain expression more than other compounds (Costa et al. 2013). Our study investigated changes in miRNA expression, using Hematopoietic stem cell isolated from cord blood before starting sodium butyrate and after reaching sodium butyrate.

Materials and methods

Cell isolation and culture

Human cord blood was obtained from Iran Blood Transfusion Organization. In the first step, cord blood was diluted by Hydroxyethyl Starch (HES) solution with the ratio of 1:7 to precipitate red blood cells. Then, mononuclear cells (MNCs) were isolated using a Hypaque-Ficoll density gradient (Amersham Pharmacia, Piscataway, NJ) with one/two ratio (ficoll to cord blood ratio). By centrifuging at 400 × g for 40 min, mononuclear layer was separated from red blood cell layer, and then it was mixed with three volumes of phosphate buffered saline (PBS) with pH = 7.2. After centrifuge at 300 × g for 10 min, pelleted mononuclear cells were washed with PBS twice.

For isolation of CD133⁺cells from other MNCs, magnetic activated cell sorting (MACS) CD133⁺isolation kit (Miltenyi Biotech, Germany) was used according to manufacturer’s instruction. Briefly, harvested MNCs were passed through MACS column placed in a magnetic field, and about 5 × 10⁵ CD133⁺cells were purified with approximately 95% purity. The isolated CD133⁺cells suspended in Iscove’s Modified Dulbecco’s Medium (IMDM) containing, 25% (v/v) fetal bovine serum (FBS) (Cambrex, Belgium), 3 U/mL erythropoietin (EPO; R&D systems, Minneapolis, MN, USA), 5 ng/ml interleukin-3 (IL-3; Stem cell Technology Vancouver, BC, Canada), and 2 mM L-glutamine at an initial density of 10⁵cells/mL for 14 days. Sodium butyrate (Sigma, Saint Louis, MO) were dissolved in distilled water to obtain a stock concentration of 500 mM. The stock solution was diluted with culture medium and added to the cells at a final concentration of 100 μM sodium butyrate (Ahmadvand et al. 2014) on the second week (day 7–14). The medium was replaced with fresh medium every 3 days. The cells were counted and viability was measured with 0.4% trypan blue dye. Isolated CD133⁺divided into two groups and treated with (1) DW, as a vehicle control and (2) sodium butyrate at a concentration of 100 μM. After 14 days, erythroid progenitor collected and RNA extraction was performed.

Flow cytometry analysis

Flow cytometry analysis was performed by MACS technique to determine the purity of isolated CD133 + cells. Monoclonal antibody against CD133 + was conjugated with PE (clone, AC141; Miltenyi Biotech, Germany), and PE conjugated-mouse IgG1 antibody (IQ-Products, the Netherlands; IQP-191F), as an isotype of negative control, were added to about 10⁵ isolated cells according to the manufacturer’s protocol. The mixture incubated for 60 min at 4 °C, and 100 μl Paraformaldehyde 1% solution was added in order to fix the cells and their interactions with florescent antibodies. Flow cytometry analysis was carried out by FACScan flow cytometer (Becton Dickinson, San Diego, CA) and the results were analyzed by FlowJo™ software.

RNA extraction

RNA was isolated from both treated and untreated cell samples to identify genes that were either activated or repressed by sodium butyrate treatment. The cells were centrifuged at 300 rpm for 5 min and the supernatant was removed. Subsequently, 200 μl of TRIZOL per 10⁶ cells was added, mixed by vortex for 15 s, and leaved at room temp for 5 min. Then, solution was transferred to 1.5 ml microtube, and 0.2 ml chloroform/1 ml TRIZOL reagent was added. Samples mixed vigorously by vortex for 1 min, incubated at 4 °C for 5 min and then centrifuged at 12000 rpm for 20 min at 4 °C. Following centrifugation, the mixture separated into three visible phases: phenol-chloroform phase (lower red), an interphase, and a colorless upper aqueous phase. RNA remains exclusively in the aqueous phase. Upper aqueous phase was transferred carefully into fresh labeled tubes without disturbing the interphase. RNA was precipitated from the aqueous phase by mixing with Ethanol (0.5 ml 100% cold Ethanol/1 ml TRIZOL Reagent). Following the overnight incubation in −70 °C, the tube was spun at 12,000 rpm for 45 min at 4 °C. Then the ethanol was aspirated and the pellet was washed with 1 ml of 70% ethanol, mixed by vortex and spun for 15 min. Afterward, the ethanol was removed and tube was incubated at room temperature for 5 minutes to allow the remaining ethanol to air dry. The pellet was resuspended in 30–50 μl RNase free water (DEPC). The isolated RNA was subjected to reverse transcription to generate cDNA molecules.

Reverse transcription PCR

cDNA synthesis was performed using TaKaRa RR037A in reaction mixture containing the component which is presented in Table 1. The reaction mixture incubated at 37 °C for 15 min (1 cycle) and at 85 °C for 5 s (1 cycle). The synthesized cDNA was stored at −20 °C for later qPCR.

Table 1. cDNA synthesis reaction mixture.

Reagents	Amount (μl)	Final concentration
5 × PrimeScript Buffer (for Real Time)	2	1×
PrimeScript RT Enzyme Mix	0.5
Oligo dT Primer (50 μM)	0.5	25 pmol
Random 6 mers (100 μM)	0.5	50 pmol
Total RNA	3
RNaseFree dH2O	3.5
Total	10

Relative quantification PCR

Quantification PCR (Q-PCR) was performed to evaluate γ-globin expression and erythroid differentiation. It was carried out using the Rotor gene 6000 Real-Time PCR system (Applied Corbett, USA) by the ΔΔCt method, with HPRT as the internal control to normalize RNA levels. Primers sequences are shown in Table 2. PCR amplification was carried out in a 20 μl PCR reaction volume containing forward primer 10 pmol (1 μm), reverse primer 10 pmol (1 μm), Master mix Eva green (10), cDNA (2), and distilled water (6). The PCR conditions were as follows: denaturation at 95 °C for 5–10 min (1 cycle), annealing at 94 °C for 15 s (40 cycles), extension at 60 °C for 40 s (40 cycles), and a final extension step at 72 °C for 20 s (40 cycles).

Table 2. Primers used for relative quantification PCR.

Reverse primer	Forward primer	Genes
5′CGGTCACCAGCACATTTCC3′	5′GTCCTCTGCCTCTGCCATC3′	γ-globin
5′GACCGAGATGGTGGAAACTG3′	5′CTTTGGACATGCTCATCTGG3′	CD71
5′GAGGTTTTACATCAGATGGGCTTT3′	5′GGCTGGTGTTATTGGAACGATC3′	CD235a
5′AATTACTTTTATGTCCCCT- GTTGACTGG3′	5′GCTATAAATTCTTTGCTGACC- TGCTG3′	HPRT*

*HPRT gene was used as a housekeeping gene for normalization of RNA levels.

Microarray analysis

Relative expression of miRNA isolated from erythroid cells derived from CD133⁺cells was measured using a human miRNA microarray (AffymetrixGeneChip® miRNA 4.0 Array, Affymetrix, Santa Clara, CA) by Macrogen Company in Korea. The Affymetrix GenechipmiRNA array process was performed according to the manufacturer's protocol. For this purpose, 1 μg RNA samples were labeled with the FlashTag™ Biotin RNA Labeling Kit (Genisphere, Hatfield, PA). The labeled RNA was quantified, fractionated and hybridized to the miRNA microarray according to the standard procedures provided by the manufacturer. Then, it was heated to 99 °C for 5 min and to 45 °C for 5 min. RNA-array hybridization was performed with agitation at 60 rotations per minute for 16–18 h at 48 °C on an Affymetrix® 450 Fluidics Station. The chips were washed and stained using a Genechip Fluidics Station 450 (Affymetrix, Santa Clara, CA), and then scanned with an AffymetrixGeneChip Scanner 3000 (Affymetrix, Santa Clara, CA). Signal values were computed using the Affymetrix® GeneChip™ Command Console software.

We run qPCR for miR-940 and miR-3663 to confirm our microarray results. The process consist of microRNA cDNA synthesizing by cDNA synthesis kit (Stratagene, La Jolla, CA) in two stages: first, polyadenylation was performed by reaction mixture presented in Table 3. The reaction mixture incubated at 37 °C for 30 min (1 cycle) and at 95 °C for 5 min (1 cycle). The second stage was performed according to Tables 4 and 5.

Table 3. miRNA cDNA synthesis reaction mixture.

Reagents	Amount (μl)
5 × poly A polymerase buffer	4.0
rATP (10 mM)	1.0
Total RNA (30 ng–1 μg)	4
E coli poly A polymerase	1
RNaseFree dH2O	3
Total	20

Table 4. miRNA qPCR reaction mixture.

Reagents	Amount (μl)
10 × AffinityScript RT buffer	2.0
The polyadenylation reaction	4
dNTP mix (100 mM)	0.8
RT adaptor primer (10 μM)	1.0
Affinity Script RT/RNase Block enzyme mixture	1.0
RNase-free water	11.2
Total	20

Table 5. cDNA synthesis temperature profile.

Time (min)	Temperature (°C)
5	55
15	25
30	42
5	95

miRNA qPCR

In order to evaluate miR-940 and miR-3663 expression, qPCR was carried out using the Rotor gene 6000 Real-Time PCR system (Applied Corbett, USA) by the ΔΔCt method, with U6 as internal control to normalize RNA levels. Primers’ sequences are shown in Table 6. PCR amplification was carried out in a 20 μl PCR reaction volume containing 3.123 μM miRNA specific forward PRIMER (1 μl), 3.125 μM universal reverse primer (1 μl), diluted reference dye (0.375 μl), Master mix Eva green (12.5 μl), cDNA (1 μl), and distilled water (4.56 μl). The PCR conditions were as follows: denaturation at 95 °C for 10 min (1 cycle), then 40 cycles of denaturation at 95 °C for 10 s, annealing at 60 °C for 15 s, and extension step at 72 °C for 20 s.

Table 6. Primers’ sequences of qPCR.

	Mir940	Mir3663
Forward	CTCTTAGGGCCTTGGGTTTG	GGACACAGCAGTCATGGCTTC
Revers	GTCACTTAGGCTGCTCAT	TACCCTCTCTGAGCCGCAATTC

Data preparation and statistical analysis

Raw data were extracted automatically in Affymetrix data extraction protocol using the software provided by AffymetrixGeneChip® Command Console® (AGCC). miRNA level RMA + DABG-All was analyzed and results were exported using Affymetrix® Expression Console™ Software. Array data were filtered by probes annotated species. The comparative analysis between test and control samples was carried out using fold-change. All statistical test and visualization of differentially expressed genes was conducted using R statistical language version 2.15.0.

Results

Isolation of CD133 + cells

CD133 + stem cells were isolated successfully from cord blood sample (Figure 1a). The purity of isolated cells was assessed using flow cytometry (Figure 1b). Samples with expression values more than 90% were selected for further studies. In microscopy analysis, each sample of isolated CD133 + hematopoietic stem cells consisted about 1 million cells. Cell viability was about 87% as counted using Trypan blue assay.

Figure 1. CD133 + stem cells isolation. (a) Cord blood-derived CD133 + stem cells in day 2 of differentiation. (b) Flowcytometry analysis of isolated CD133 + indicates 97% purity on the day of isolation.

RNA extraction

On day 14, 10⁶ CD133 + cells were isolated from culture medium for RNA extraction. Subsequently, RNA samples were run on a 1% agarose gel by electrophoresis to determine the purity and probable degradation of the extracted RNA. As presented in Figure 2(a), in lane 4 and 3, bands can be observed which are the three RNA subunits 28S, 18S, and 5S rRNA. For quality control of the samples (control and test), A260/280 absorbance ratio and RNA concentration was measured by NanoDrop. A260/280 absorbance ratio and RNA concentration were found to be 2.25, 217.24 ng/μl and 2.24, 188.34 ng/μl for control and test, respectively. Also for quality control of the RNA samples, RNA Integrity Number (RIN) assessed using Agilent Technologies 2100 Bioanalyzer in test (Figure 2b) and control (Figure 2c) samples, and the results reported based on Integrity Number ≥ 8. RIN was estimated at 8 for control and 8.5 for test samples.

Figure 2. Electrophoresis of extracted RNA on day 0 and 14 of differentiation. 18S rRNA band (800 bp) and 28S rRNA band (1500 bp) of control and test samples on days 0 and 14 of differentiation are shown in the picture (a). RNA Integrity Number (RIN) in test (HSC-SB) (b) and control (HSC-C) (c) samples. (1) Pre-region marker; (2) 5s region; (3) fast region; (4) 18s fragment; (5) inter region; (6) 28s fragment; (7) precursor region; (8) post-region.

Erythroid differentiation of CD133 + cells

In order to evaluate the effects of sodium butyrate on the induction of gamma globins gene expression in erythroid precursors, CD133 + cells were cultured in IMDM medium containing 100 μM sodium butyrate, for a period of 8 days. Subsequently, the expression of erythroid markers including CD71, CD235a and gamma globins was determined using qPCR technique. Fold changes in expression of these genes are presented in Figure 3.

Figure 3. Relative expression of CD71, CD235a and Gamma globin in differentiated cells in comparison with control cells on days 0 and 14. Relative expression of CD71, CD235a, and Gamma globin significantly elevated in differentiated cells in comparison with control cells. Errors bars represent standard deviation. *P = 0.001 and **P = 0.0001.

Quality control of miRNAs

Quality control of the miRNAs was performed by Macrogen Company to ensure the quality of miRNAs for microarray procedure. microRNAs concentrations and their attachment to probe were acceptable (6). Results are presented in Figure 4.

Figure 4. Quality control of the miRNAs. (a) The probes which attached to each sample. The probes show acceptable signals and weak signals (Flag A) were omitted; (b) Concentration graphs and probes distribution after normalization. Concentration graphs show normalized values and others indicate biases and statistical errors; (c) Distribution graphs and comparison between samples by Pearson product-moment correlation coefficient (Range: −1 ≤ r ≤ 1). This graph shows normalized values. X axis indicates control sample (HSC-C) and Y axis indicates test sample (HSC-SB).

miRNAs expression

Microarray assay was performed by Macrogen Company. In order to minimize systematic errors and to avoid false statistical results, rare data were filtered and normalized. Results are demonstrated in Figure 5 and Table 7.

Table 7. Altered microRNAs and their related chromosome, sequence lengths, and sequences.

Transcript ID (array design)	HSC-SB/HSC-C.fc	Sequence type	Alignments	Sequence length	Sequences
hsa-let-7b-5p	−2.211425	miRNA	chr22:46509571-46509592 (+)	22	UGAGGUAGUAGGUUGUGUGGUU
hsa-let-7f-5p	−7.607661	miRNA	chr9:96938635-96938656 (+)//chrX:53584207-53584228 (−)	22	UGAGGUAGUAGAUUGUAUAGUU
hsa-let-7f-1-3p	2.55333	miRNA	chr9:96938691-96938712 (+)	22	CUAUACAAUCUAUUGCCUUCCC
hsa-miR-21-5p	−2.473677	miRNA	chr17:57918634-57918655 (+)	22	UAGCUUAUCAGACUGAUGUUGA
hsa-miR-30a-5p	−4.805625	miRNA	chr6:72113298-72113319 (−)	22	UGUAAACAUCCUCGACUGGAAG
hsa-miR-199a-3p	−2.307177	miRNA	chr1:172113694-172113715 (−)//chr19:10928105-10928126 (−)	22	ACAGUAGUCUGCACAUUGGUUA
hsa-miR-7-5p	−7.502118	miRNA	chr15:89155087-89155109 (+)//chr19:4770712-4770734 (+)//chr9:86584727-86584749 (−)	23	UGGAAGACUAGUGAUUUUGUUGU
hsa-miR-181a-2-3p	2.499995	miRNA	chr9:127454797-127454818 (+)	22	ACCACUGACCGUUGACUGUACC
hsa-miR-199b-3p	−2.307177	miRNA	chr9:131007024-131007045 (−)	22	ACAGUAGUCUGCACAUUGGUUA
hsa-miR-204-3p	−2.770119	miRNA	chr9:73424909-73424929 (−)	21	GCUGGGAAGGCAAAGGGACGU
hsa-miR-221-3p	−2.560584	miRNA	chrX:45605608-45605630 (−)	23	AGCUACAUUGUCUGCUGGGUUUC
hsa-let-7g-5p	−3.289816	miRNA	chr3:52302352-52302373 (−)	22	UGAGGUAGUAGUUUGUACAGUU
hsa-miR-27b-5p	2.449231	miRNA	chr9:97847745-97847766 (+)	22	AGAGCUUAGCUGAUUGGUGAAC
hsa-miR-130a-3p	−5.439813	miRNA	chr11:57408725-57408746 (+)	22	CAGUGCAAUGUUAAAAGGGCAU
hsa-miR-140-5p	−2.742847	miRNA	chr16:69967006-69967027 (+)	22	CAGUGGUUUUACCCUAUGGUAG
hsa-miR-152-3p	−2.354822	miRNA	chr17:46114540-46114560 (−)	21	UCAGUGCAUGACAGAACUUGG
hsa-miR-138-1-3p	2.460313	miRNA	chr3:44155766-44155787 (+)	22	GCUACUUCACAACACCAGGGCC
hsa-miR-34b-3p	2.180183	miRNA	chr11:111383712-111383733 (+)	22	CAAUCACUAACUCCACUGCCAU
hsa-miR-301a-3p	−6.909709	miRNA	chr17:57228510-57228532 (−)	23	CAGUGCAAUAGUAUUGUCAAAGC
hsa-miR-130b-5p	−2.346938	miRNA	chr22:22007605-22007625 (+)	21	ACUCUUUCCCUGUUGCACUAC
hsa-miR-365b-5p	−2.210533	miRNA	chr17:29902458-29902479 (+)	22	AGGGACUUUCAGGGGCAGCUGU
hsa-miR-451a	−2.799991	miRNA	chr17:27188421-27188442 (−)	22	AAACCGUUACCAUUACUGAGUU
hsa-miR-410-3p	−2.581299	miRNA	chr14:101532298-101532318 (+)	21	AAUAUAACACAGAUGGCCUGU
hsa-miR-489-3p	2.121202	miRNA	chr7:93113259-93113280 (−)	22	GUGACAUCACAUAUACGGCAGC
hsa-miR-493-5p	−2.437498	miRNA	chr14:101335412-101335433 (+)	22	UUGUACAUGGUAGGCUUUCAUU
hsa-miR-181d-5p	−2.59999	miRNA	chr19:13985724-13985746 (+)	23	AACAUUCAUUGUUGUCGGUGGGU
hsa-miR-503-5p	−2.346938	miRNA	chrX:133680401-133680423 (−)	23	UAGCAGCGGGAACAGUUCUGCAG
hsa-miR-595	−3.378137	miRNA	chr7:158325425-158325445 (−)	21	GAAGUGUGCCGUGGUGUGUCU
hsa-miR-454-3p	−2.459462	miRNA	chr17:57215148-57215170 (−)	23	UAGUGCAAUAUUGCUUAUAGGGU
hsa-miR-670-5p	−2.602234	miRNA	chr11:43581225-43581244 (+)	20	GUCCCUGAGUGUAUGUGGUG
hsa-miR-675-3p	−2.578938	miRNA	chr11:2017998-2018017 (−)	20	CUGUAUGCCCUCACCGCUCA
hsa-miR-374b-5p	−3.835062	miRNA	chrX:73438422-73438443 (−)	22	AUAUAAUACAACCUGCUAAGUG
hsa-miR-301b	−2.433797	miRNA	chr22:22007314-22007336 (+)	23	CAGUGCAAUGAUAUUGUCAAAGC
hsa-miR-940	2.106382	miRNA	chr16:2321807-2321827 (+)	21	AAGGCAGGGCCCCCGCUCCCC
hsa-miR-1290	−2.942842	miRNA	chr1:19223572-19223590 (−)	19	UGGAUUUUUGGAUCAGGGA
hsa-miR-1255b-5p	2.032245	miRNA	chr1:167967903-167967924 (+)//chr4:36428027-36428048 (−)	22	CGGAUGAGCAAAGAAAGUGGUU
hsa-miR-2276-3p	3.306113	miRNA	chr13:24736608-24736629 (+)	22	UCUGCAAGUGUCAGAGGCGAGG
hsa-miR-3124-5p	−3.422843	miRNA	chr1:249120582-249120602 (+)	21	UUCGCGGGCGAAGGCAAAGUC
hsa-miR-3157-3p	−5.015078	miRNA	chr10:97824078-97824099 (−)	22	CUGCCCUAGUCUAGCUGAAGCU
hsa-miR-3193	−3.923687	miRNA	chr20:30194989-30195010 (+)	22	UCCUGCGUAGGAUCUGAGGAGU
hsa-miR-3201	2.739582	miRNA	chr22:48670176-48670192 (+)	17	GGGAUAUGAAGAAAAAU
hsa-miR-4290	−2.163305	miRNA	chr9:92785733-92785751 (−)	19	UGCCCUCCUUUCUUCCCUC
hsa-miR-3613-5p	3.79172	miRNA	chr13:50570601-50570622 (−)	22	UGUUGUACUUUUUUUUUUGUUC
hsa-miR-3663-3p	2.099997	miRNA	chr10:118927206-118927228 (−)	23	UGAGCACCACACAGGCCGGGCGC
hsa-miR-378e	−2.136297	miRNA	chr5:169455549-169455567 (+)	19	ACUGGACUUGGAGUCAGGA
hsa-miR-378h	−3.466308	miRNA	chr5:154209026-154209046 (+)	21	ACUGGACUUGGUGUCAGAUGG
hsa-miR-4470	2.923507	miRNA	chr8:62627391-62627411 (+)	21	UGGCAAACGUGGAAGCCGAGA
hsa-miR-4684-3p	−2.258048	miRNA	chr1:23046059-23046080 (+)	22	UGUUGCAAGUCGGUGGAGACGU
hsa-miR-4708-3p	−10.744661	miRNA	chr14:65801841-65801862 (−)	22	AGCAAGGCGGCAUCUCUCUGAU
hsa-miR-4726-5p	−2.134967	miRNA	chr17:36875944-36875966 (+)	23	AGGGCCAGAGGAGCCUGGAGUGG
hsa-miR-4778-5p	3.766815	miRNA	chr2:66585429-66585450 (−)	22	AAUUCUGUAAAGGAAGAAGAGG
hsa-miR-5572	−2.233918	miRNA	chr15:80873520-80873540 (+)	21	GUUGGGGUGCAGGGGUCUGCU
hsa-miR-6080	−2.581299	miRNA	chr17:62776914-62776932 (+)	19	UCUAGUGCGGGCGUUCCCG
hsa-miR-6083	2.061385	miRNA	chr3:124093255-124093274 (+)	20	CUUAUAUCAGAGGCUGUGGG
hsa-miR-6084	2.008771	miRNA	chr1:20960249-20960268 (+)	20	UUCCGCCAGUCGGUGGCCGG
hsa-miR-6721-5p	−2.59999	miRNA	chr6:32137861-32137883 (−)	23	UGGGCAGGGGCUUAUUGUAGGAG
hsa-miR-6737-5p	2.297187	miRNA	chr1:153934871-153934891 (−)	21	UUGGGGUGGUCGGCCCUGGAG
hsa-miR-6747-3p	−4.365019	miRNA	chr11:62334483-62334503 (−)	21	UCCUGCCUUCCUCUGCACCAG
hsa-miR-6780b-5p	−2.194173	miRNA	chr6:43402285-43402307 (+)	23	UGGGGAAGGCUUGGCAGGGAAGA
hsa-miR-6837-5p	−2.190419	miRNA	chr7:44091370-44091390 (+)	21	ACCAGGGCCAGCAGGGAAUGU
hsa-miR-6885-3p	−2.192163	miRNA	chr19:6389649-6389669 (−)	21	CUUUGCUUCCUGCUCCCCUAG
hsa-miR-7109-5p	−2.082968	miRNA	chr22:32017492-32017512 (−)	21	CUGGGGGGAGGAGACCCUGCU
hsa-miR-7111-5p	−2.042549	miRNA	chr6:35438286-35438307 (+)	22	UGGGGGAGGAAGGACAGGCCAU
hsa-miR-7152-3p	2.019997	miRNA	chr10:73550538-73550557 (+)	20	UCUGGUCCUGGACAGGAGGC
ACA17	−2.136297	HAcaBox	chr9:139621199-139621331 (−)	133	ACTGCCCCTAGAGGCGTTGCAGCTGTGGCTGCCGTGTCACATCTGTGTCATTAGGTGGCAGAGATTAGAGAGGCTATGTCTACGCTCAGCGTTCTGCCCCGTGAACGTTTGAATGTTTGATAGTCTCACACTC
ACA19	−2.034312	HAcaBox	chr10:120819523-120819650 (−)	128	GTGCACATTTCATTGACCTGCTTTCTTTTATGTGAGTAGTGTTATTTCTTATGTGCTATACAAATAATTGAAGGCTAATTAGCAGTATAACTATAAATAGTAATGCTGCCAGTCTCCTTCAGACAAAA
ACA24	−2.31683	HAcaBox	chr4:119200345-119200475 (+)	131	CTCCATGTATCTTTGGGACCTGTCAGCCGTGGCAGTCTCCCTTCCTAGCCATGGAAGAGCATATCCTTGTTTATTGGCAAAGCTGTCACCATTTAATTGGTATCAGATTCTGACTTGCACAAGTAACATTC
ENSG00000206903	−2.31683	snoRNA	chr15:65577799-65577929 (−)	131	CTCCATGTATCTTTGGGACCTGTCAAGTGTGGCAGTCTCCCTTCCTTGCCATGGAAGAGCATATTCTTGTTTACCAGCAAAGCTGTCACCATTTAATTGGTATCAGATTCTGACTTGCACAAGTAACATTC
ENSG00000207118	−2.100007	snoRNA	chr11:122929617-122929703 (−)	87	TCGCTATGATGATGGATTCCAAAACCATTCGTAGTTTCCACCAGAAAGTCTTATGTTGGCCAGTTCCTTCCTTGGATGTTTGAGCGA
ENSG00000207130	−2.31683	snoRNA	chr3:128433414-128433548 (−)	135	CTCCATGTGTCTTTGGAACCTGTCAGCTGTGGCAGTTGCCCTTCCTAGCCATGGAAGAGTAAGTATATTCTTGTTTATTGGCAAAGCTGTCACCATTTCATTGGTATCAGATTCTGACTTGCACAAGTAACATTC
HBII-296B	−2.316237	CDBox	chr17:2232419-2232504 (−)	86	AAGAGCCAATGATGTTTTTATTCAAAATGTCTGAACCTGTCTGAAGCATCCCAGTGATGCAACTTCTGTGTGATACTGAGGCTTTT
hsa-mir-1260a	2.333313	stem-loop	chr14:77732561-77732633 (+)	73	ACCUUUCCAGCUCAUCCCACCUCUGCCACCAAAACACUCAUCGCGGGGUCAGAGGGAGUGCCAAAAAAGGUAA
hsa-mir-7515	3.542559	stem-loop	chr2:6790505-6790571 (+)	67	CUUUUAUCAUUUCAUUUAUGCCUGUUAGCAUGAAAAAAGAGGCUGAGAAGGGAAGAUGGUGACAAGG
U101	−2.194766	CDBox	chr6:133136446-133136518 (+)	73	GTTTGAATGATGACTTTAATTGTCGGATACCCCTTCACTCCTTTTATGAGTGAAACATAAGAGTCTGACAAAC
U62A	−2.004627	CDBox	chr9:134361052-134361137 (+)	86	TCTCAGTGATGTAATTCCAATAGATCCTTCTGACCCTCCACTGTGGACTCAATAGCAGGGAGATGAAGAGGACAGTGACTGAGAGA
U62B	−2.004627	CDBox	chr9:134365873-134365958 (+)	86	TCTCAGTGATGTAATTCCAATAGATCCTTCTGACCCTCCACTGTGGACTCAATAGCAGGGAGATGAAGAGGACAGTGACTGAGAGA

qPCR

Our data shows a 1.8- and 1.56-fold over-expression of miR-940 and miR-3663, respectively. Results merely indicate sodium butyrate can up-regulate these microRNA (Figure 6).

Figure 5. Microarray assay. (a) The graph of miRNAs expression in test (T, red, right column) and control (C, blue, left column) groups; (b) the graph of significant changes in miRNAs expression. 289 miRNAs down-regulated or up-regulated by |fc| ≥ 1.5 and 76 miRNAs down-regulated or up-regulated by |fc| ≥ 2 (upper column upregulated, lower column downregulated); (c) the graph of expression values distribution. X axis: Control and Y axis: Test; (d) the graph of the log changes. Volume = Squared root (normalized value of Control * normalized value of Test) Red dot (marked dots): top five ranking miRNAs in volume with significant cut-off. 180 × 147 mm (300 × 300 DPI).

Figure 6. qPCR results by fold. Errors bars represent standard deviation. *P = 0.001.

Discussion

Sodium butyrate has been considered as a low-risk and highly effective agent in γ-globin gene induction (Weinberg et al. 2005). Long-term cytotoxic potential of fetal hemoglobin gene inducing agents leads to investigations into safe molecules, which regulate HbF expression and switching (Bank 2006). One of these molecules could be micro RNAs (miRNAs), which are suggested to be new approach in targeting HbF induction (Azzouzi et al. 2011). In order to find that which miRNAs are appropriate candidate for HbF induction, it is necessary to recognize that which of them are up-regulated in HbF switching. Regarding to this, sodium butyrate was used for β-chain switching to γ-chain, and then microarray method was performed to explore that which miRNAs up-regulated in this process.

In this study, significant increase was observed in γ-globin expression in sodium butyrate group. Our findings were in agreement with researches of Perrine et al. (1989, 1993), Constantoulakis et al. (1989), and Fibach et al. (1993). Sodium butyrate induced γ-gene expression in CD133 + in sodium butyrate group in comparison with the control group. Our results were in coordination with Dehghanifard et al. (2011, 2012), Ahmadvand et al. (2011), and Fard et al. (2013a,b). It has been shown that sodium butyrate is effective on SCD patients even on the ones that are resistant to HU (Panja and Basu 2015). None of these researchers evaluated microRNA expression pattern in their study. In our study, microarray analysis was performed to evaluate the miRNAs up-regulation and down-regulation pattern.

Walker et al. (2011) investigated miRNAs expression profile in erythroid cells of sickle cell anemia patient, before and after hydroxy urea administration. Findings indicated that the expression profile of some of the miRNAs was different in control group (healthy individuals), sickle cell anemia patients, and patients who treated with hydroxy urea. They suggested that hsa-mir 26b and miR151–3p are involved in HbF expression induction (Walker et al. 2011). In the study of Alijani et al. (2014), the effect of upregulation of miR-26b in the induction of HbF expression on K-562 was evaluated, and the results showed increase in HbF expression in miR-26b upregulation. Walker et al. (2011) showed that hydroxy urea leads to up-regulation in miR26b and HbF expression. Findings of Walker et al. and Alijani et al. were in accordance with each other.

In this study, miRNAs expression profile was determined by miRNA microarray technique, before and after treatment with sodium butyrate. According to our results, the expression of 76 miRNAs had significant difference in test group compared to the control group; in which, 20 of them were up-regulated, whereas 56 of them were down-regulated in the presence of sodium butyrate. As presented in Figure 5d and Table 7, miRNA-U101, hsa-miR-4726–5p, hsa-miR7109–5p, hsa-miR3663, and hsa-miR940 showed the most alterations in all miRNAs which the first three were down-regulated and the last two were up-regulated. Moreover, our findings illustrated that miR221 expression was down-regulated by about 2-fold in the test group, which treated with sodium butyrate. This was in agreement with Gabbianelli study. Gabbianelli et al. demonstrated its role in HbF production. miR221/222 down-regulate KIT expression in normal erythropoiesis. KIT activation inhibits HbF production; therefore, down-regulation of KIT expression leads to increase in HbF production (Gabbianelli et al. 2010).

It should be noted that HbF inducers can act only as a treatment for hemoglobinopathies to ameliorate the clinical condition of patients along with blood transfusion. However, gene therapy could be a definitive cure. Hence identification of microRNAs involve in the gamma chain expression can lead to novel therapeutic approaches in hemoglobinopathies. We suggest future studies to focus on determining the role of five demonstrated miRs in epigenetic modification, so it would illuminate the way to find a cure for beta thalassemia.

Conclusion

Our results in the current study confirmed the sodium butyrate as an inducer of HbF. However, due to the complications of HbF inducers, it seems crucial to find a substitute for these products; miRNAs can be the promising surrogate. We found that some miRNAs were up-regulated in the sodium butyrate presence; therefore, they can be used for inducing γ-globin gene expression in β-thalassemia and sickle cell anemia patients.

Funding information

This study was supported by grant from the Mashhad University of medical sciences (research project number 910736).

Acknowledgements

This study was done in Department of Hematology of Tehran University of medical sciences. The authors thank all authorities and office worker of faculty of hematology of Tehran and Mashhad University of medical sciences for their kind assistance.

Disclosure statement

The authors report no conflicts of interest.