Dust mites have been determined to be the predominant allergen in asthma (7). In the present study, we focused on dust mite-induced asthma. The study was approved by the Bioethics Committee of the Shanghai Children's Hospital (Shanghai, China), and written informed consent was obtained from the participants or their guardians.

A total of 62 pediatric patients with asthma were recruited, as the case group, from our Asthma Clinic at Shanghai Children's Hospital (Shanghai, China). A diagnosis of asthma was made according to the Global Initiative for Asthma (GINA) guidelines (10). We focused on patients with intermittent-mild asthma in this study, as the pathogenesis of severe asthma is more complex (30). Detailed information regarding patient demographics and enrollment inclusion and/or exclusion criteria is provided in Table I. Forty-one boys and 21 girls (average age, 6.05±2.15 years) were recruited for the asthma case group. During the time that they were enrolled in this study and their blood specimens were being collected, these patients did not have current or recent infections, were not taking any medication, such as inhaled corticosteroid (ICS) (31), and did not experience any acute asthma exacerbations or attacks, in order to avoid some disturbing factors. An independent, unrelated group of 37 boys and 25 girls (average age, 7.83±3.30 years) showing no clincial phenotypes was enrolled as normal controls. No significant difference was observed between the groups of cases and controls, in terms of gender and age (χ²=0.667, p=0.414).

Total IgE and specific serum IgE, which were thresholds of enrollment inclusion (Table I), were assessed using the ImmunoCAP System (Thermo Fisher Scientific/Phadia AB, Uppsala, Sweden). The cut-off of total IgE was 60 kU/l. In the asthma group, total IgE was >60 kU/l, and IgE serology atopic sensitization was indicated if the child had only dust mite allergen-specific serum IgE (≥0.35 kU/l) (data not shown) without food allergens of Fx5 (egg white, milk, fish, wheat, peanut and soy) (32). In the control group, no allergen-specific IgE was >0.35 kU/l, and total IgE was <60 kU/l. The percentage of eosinophils in peripheral blood (EOS) was also counted in order to confirm the patient's atopy. The cut-off value was 4%.

Twelve pairs of gender- and age-matched children with asthma and normal control individuals were subjected to initial microarray-based discovery analysis. Peripheral blood (5 ml) was collected from each participant using ethylenediaminetetraacetic acid (EDTA) as the anticoagulant. Total RNA was isolated using TRIzol (Invitrogen, Grand Island, NY, USA) and an miRNeasy mini kit (Qiagen, Valencia, CA, USA) kits according to the manufacturer's instructions, which efficiently recovered all RNA species, including miRNAs. The quality and quantity of RNAs collected from all participants were measured using a NanoDrop spectrophotometer ND-1000 (NanoDrop Technologies, Wilmington, DE, USA), and RNA integrity was determined by gel electrophoresis. After RNA was isolated from the samples, the miRCURY LNA™ microRNA Hy3™/Hy5™ Power labeling kit (Exiqon, Vedbaek, Denmark) was used according to the manufacturer's instructions for miRNA labeling. Briefly, 1 μg of each sample was 3′-end-labeled with Hy3 fluorescent dye, using T4 RNA ligase according to the following procedure: RNA in 2.0 μl of water was combined with 1.0 μl CIP buffer and CIP (Exiqon). The mixture was incubated for 30 min at 37°C and was terminated at 95°C for 5 min. Then, 3.0 μl labeling buffer, 1.5 μl fluorescent label (Hy3), 2.0 μl dimethyl sulfoxide (DMSO), and 2.0 μl labeling enzyme were added into the mixture. The labeling reaction was incubated for 1 h at 16°C and was terminated by incubation at 65°C for 15 min. After termination, the Hy3-labeled samples were hybridized on the miRCURY LNA array (v.18.0; Exiqon) according to the manufacturer's instructions. The total 25 μl mixture from Hy3-labeled samples together with 25 μl hybridization buffer were denatured at 95°C for 2 min, incubated in ice for 2 min, and then hybridized to the microarray at 56°C for 16–20 h with a 12-Bay Hybridization system (Nimblegen Systems, Inc., Madison, WI, USA). After hybridization, the slides were prepared, washed using a wash buffer kit (Exiqon), and dried by centrifugation at 200 × g for 5 min. The slides were then scanned using the Axon GenePix 4000B microarray scanner (Axon Instruments, Union City, CA, USA).

The signal quantification of scanned microarray images was captured using GenePix Pro 6.0 software (Molecular Devices, Sunnyvale, CA, USA). Replicated miRNAs were averaged, and miRNAs whose intensities were ≥30 in all samples were selected for calculation of the normalization factor. Expressed data were normalized using the median normalization. Significant differentially expressed miRNAs were identified through volcano plot filtering. Hierarchical clustering was performed to show distinguishable miRNA expression profiling among samples using MEV software (v4.6, TIGR; www.tm4.org/). A threshold (fold-change ≥2.0 and p-value ≤0.05) was used to determine the significance of differences of up- or downregulated miRNAs. The heat map diagram shows the result of the two-way hierarchical clustering of miRNAs and samples.

Web-based programs, namely miRBase (http://mirbase.org/), PicTar (http://pictar.mdc-berlin.de/), miRDB (http://mirdb.org/miRDB/), miRanda (http://www.microrna.org/microrna/) and TargetScan (http://www.targetscan.org/), were employed to predict miRNA target(s). The miRNA-targeted transcripts, predicted by a minimum of three programs as the miRNA-targeted candidates, were selected for further analysis. The functional enrichment was analyzed using the DAVID program (http://david.abcc.ncifcrf.gov/), in which gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were provided for analysis. Venny software (http://bioinfogp.cnb.csic.es/tools/venny/) was used to show the predicted target genes that are co-regulated by multi-miRNA.

To validate the microarray results, miRNA expression was quantified in all 124 samples using RT-qPCR. The primers used for reverse transcription are presented in Table II, and the PCR conditions are shown in Table III. miRNA-U6 (33) was applied as an internal control. Briefly, reverse transcription was performed using MMLV Reverse Transcriptase and RNase inhibitor (both from Epicentre, Madison, WI, USA), 10X buffer (250 mM Tris-HCl, pH 8.3; 200 mM KCl; 40 mM MgCl₂; 5 mM DTT), 2.5 mM dNTP with RT primers at 16°C for 30 min, followed by 42°C for 40 min and 85°C for 5 min. Twenty microliters of RNA (final concentration 50 ng/μl) was used as a template. The cDNA products from reverse transcription reactions were quantified using a real-time PCR System 7500 (Applied Biosystems, Foster City, CA, USA). The resulting amplicon was quantified by RT-qPCR with end-point SYBR-Green fluorescence. Each assay was performed in triplicate. RT-qPCR was also performed to determine the expression of the target genes regulated by the miRNAs that were discovered in our microarray analysis. The reverse transcription reaction was performed as descibed above except using SuperScript™ III reverse transcriptase (Invitrogen). RT primers were added into the reaction mixture at 37°C for 1 min, followed by 50°C for 60 min and 75°C for 15 min. Twenty microliters of RNA (final concentration 75 ng/μl) was used as a template. The cDNA products from the reverse transcription reactions were quantified using a real time PCR system 7500 (Applied Biosystems). The resulting amplicon was quantified by RT detection of end-point SYBR-Green fluorescence. Each assay was performed in triplicate. The primers and reaction conditions are presented in Table IV. The relevant quantification was obtained for each sample with normalization to the internal control β-actin.

The plasma concentrations of IL-4, -6, -10, -12 and -13; γ-IFN; and tumor necrosis factor-α (TNF-α) were determined using an enzyme-linked immunosorbent assay (ELISA) kit (Multiscience Biotech Co. Ltd., Shanghai, China).

The differences between the groups were analyzed using the Student's t-test when two groups were compared or by one-way ANOVA when > two groups were compared. Analyses were performed using SPSS 19.0 software (IBM, Armonk, NY, USA). A p-value <0.05 was considered to indicate a statistically significant difference.

A total of 122 differentially expressed miRNAs were identified from 12 children with asthma in a discovery study with a micro-array that detects 3,100 genome miRNAs. Among differentially expressed miRNAs, 112 were identified as upregulated, and 10 were downregulated (Fig. 1A and B). To confirm the micro-array results, five upregulated (miR-let-7d-3p, miR-550b-3p, miR-501-5p, miR-4312 and miR-675-3p) and five downregulated (miR-126-3p, miR-22-3p, miR-513a-5p, miR-151a-5p and miR-625-5p) miRNAs were randomly subjected to RT-qPCR with the 12 pairs of RNA samples that were subjected to the original array-based discovery study. The differential expression of four upregulated (miR-let-7d-3p, miR-550b-3p, miR-501-5p and miR-675-3p) and all five downregulated miRNAs were validated to be significant (p<0.05) (Table V). Three downregulated miRNAs (miR-625-5p with 2.17-fold decrease, miR-513a-5p with 2.07-fold-change, and miR-22-3p with 2.21-fold-change) (Table VI), which were statistically significant differences (p<0.05) and functionally linked to inflammation and apoptosis by bioinformatics analysis (Fig. 2), were applied for prediction of targeting transcripts.

To further confirm the microarray results, an additional independent 100 RNA samples, including 50 from individuals with asthma and 50 from control subjects, were subjected to RT-qPCR in order to validate the expression of miR-625-5p, miR-22-3p and miR-513a-5p. These three miRNAs were selected on the basis of predicting their targeted mRNAs: CBL, peroxisome proliferator-activated receptor gamma, coactivator 1 beta (PPARGC1B) and ESR1, which is involved in the PI3K-AKT and NF-ĸB pathways (18,19). Our results showed that the average levels of miR-625-5p, miR-22-3p and miR-513a-5p in the asthma group were significantly lower than those in the control group (1.13±0.22 vs. 2.21±0.17, 0.84±0.17 vs. 1.58±0.27 and 1.03±0.14 vs. 2.29±0.26, respectively; p<0.01) (Fig. 3A). No statistical significance was observed, regarding the expression levels of these three miRNAs, among individual participants within either the asthma group or the control group (p>0.05).

A total of 418 genes were predicted as the targets of miRNA-513a-5p, 257 targeted by miRNA-22-3p, and 347 targeted by miRNA-625-5p, respectively, in three web-based programs (miRDB, miRBase and miRanda). These listed target genes were subjected to further analysis with Venny software, through which miRNA targets were found to be overlapped (Fig. 1C). Three mRNAs (PLCXD3, MSL2 and BRWD3) are the overlapped targets of miRNA-513a-5p, miRNA-22-3p and miRNA-625-5p. Three pairs of any two miRNAs combined may have 12 target genes co-regulated. However, with bioinformatics functional analysis using the GO and KEGG pathways, only three transcripts of mRNAs, CBL, PPARGC1B and ESR1, were predicted as the targets of miR-513a-5p, miR-22-3p and miR-625-5p, respectively (34). Transcripts of CBL and PPARGC1B were dually targeted by miR-513a-5p and miR-22-3p, and that of ESR1 was targeted by miR-625-5p and miR-22-3p. mRNAs of CBL, PPARGC1B and ESR1 were differentially elevated in the asthma group (CBL, 2.61E-02 vs. 1.66E-02; PPARGC1B, 3.42E-02 vs. 1.94E-02; ESR1, 6.28E-03 vs. 3.67E-03; p<0.01), as compared with the control group (Fig. 3B) and have been found to be associated with the PI3K-AKT and NF-ĸB signaling pathways (Fig. 4) (35,36). It was also confirmed by GO and KEGG pathway analysis in DAVID software that the CBL was significantly associated with T- and B-cell differentiation and inflammatory factor signaling pathways (GO enrichment value >1.9, p<0.05).

Our assessment showed that in the asthma group, levels of EGFR declined (1.26E-03 vs. 2.71E-03; p<0.01), whereas levels of the mRNA of SYK involved in this pathway were elevated (4.30E-02 vs. 3.05E-02; p<0.01) (Fig. 3B).

We further examined the concentrations of inflammatory cytokine factors (γ-IFN, TNF-α, IL-13, IL-12, IL-4, IL-6 and IL-10) which may be associated with the Th1, Th2 and inflammation pathways PI3K-1KT and NF-κB that are mediated by the miRNA-targeted genes CBL, PPARGC1B and ESR1 (18,19,37) (Table VII). Decreased concentrations of γ-IFN, TNF-α, IL-12 and IL-10 in the plasma were found in the asthma group, compared with those in the control group (3.44±2.87 vs. 7.08±4.30 pg/ml, 93.80±83.41 vs. 221.06±150.33 pg/ml, 1.50±0.66 vs. 3.43±1.37 pg/ml, and 2.39±2.47 vs. 3.74±1.98 pg/ml, respectively, p<0.05). These decreased concentrations may be associated with the different expression of upstream genes with a regulatory function. The cytokine concentrations of IL-4, IL-6 and IL-13 in the asthma group were neither changed nor showing statistically significant differences compared with those in the control group (p>0.05).