Introduction http://www.sanger.ac.uk/Software/Rfam/mirna/index.shtml 1 2 4 5 6 9 7 Materials and methods Samples and RNA extraction Two fetal abortuses [12 weeks (G12w) and 24 weeks (G24w) gestational age] tissues were obtained from the National Infrastructure Program of the Chinese Genetic Resources after obtaining informed consent. The G12w tissues were liver, kidney, cerebrum, lung, and heart, and the G24w tissues were the same, plus ovary, spleen, hypothalamus, pancreas, and cervical, thoracic, lumbar, and sacral spinal cord. Total RNA was isolated using TRIZOL reagent (Invitrogen) according to the manufacturer’s standard protocol, and mRNA was purified by oligotex (Qiagen). Microarray design 7 Arabidopsis thaliana Microarray fabrication 2 Sample labeling The cDNA was labeled during first strand synthesis by using a fluorescence-tagged (Cy5) random octameric primer. Briefly, 10 μg total RNA and 1 μg random primer were mixed and incubated at 70°C for 10 min, then 5 × first-strand buffer, 0.1 M DTT, 5 mM unlabeled dNTP mix, Cy5-dCTP, RNA inhibitor, and Superscript II (200 U/μl) were added and incubated at 42°C for 2 h, then denatured at 70°C for 10 min. NaOH was added to hydrolyze RNA to stop the reaction of reverse transcription, and then HEPES was added to neutralize it. The labeled cDNA was purified by using the QIAquick Nucleotide Removal Kit (Qiagene). Microarray hybridization The labeled cDNAs and Cy3-tagged oligonucleotides complementary to the control probes were dissolved in 6 × SSPE/5 × Denhardt hybridization buffer and were hybridized with the miRNA oligonucleotide microarray for 16 h at 42°C. Then the slides were washed with buffer I (2 × SSC/0.5% SDS) for 15 min at 42°C, buffer II (1 × SSC/0.1% SDS) for 10 min at 42°C, buffer III (0.1 × SSC) for 5 min at room temperature, dipped in double-distilled water for 1 min at room temperature, and then dried. The slides were scanned by an Agilent scanner (G2565AA) at 535 and 635 nm. Statistical analysis The images were split into two, Cy3 and Cy5 channels, and each channel was imported into the Imagene Software 7.0 to read the signal value. The Cy3 signal was used as reference for the spot size of each miRNA oligo on the slides. The expression level of each miRNA in the sample labeled by Cy5 was normalized by a median method according to the Cy3 signal between two microarrays. So, the Cy5 signal, after normalization, gave the expression level of each miRNA. Clustering was carried out by Genespring Software 8.0 according to Cy5 intensity. Northern blot analysis 32 10 Results MiRNA oligo nucleotide microarray construction http://microrna.sanger.ac.uk/cgi-bin/sequences/browse.pl 1 1 1 1 Fig. 1 a b c d  Distinct miRNA profiling in different organs during human development 2 Fig. 2 a b c bar blue line left C Hr K Li Lu O P S Vc Vl Vs Vt  2 2 CNS specifically and higher expressed microRNAs 3 Fig. 3 2 left  4 Fig. 4 a b 2  11 5 Fig. 5 a b left 2 miR-17-oligo miR-20-oligo miR-92-1-oligo  Discussion 12 The miRNA oligonucleotide microarray provides a novel method to carry out genome-wide microRNA profiling in human samples. We used total RNA as the sample for the microarray test, not just labeling filtered low molecular weight RNA, which could change the ratio of pre-miRNA to miRNA. So the profile we generated was that of pre-miRNAs and miRNAs. Owing to its high throughput and small sample requirement, the miRNA OMA can be used as screening method in miRNA research. 13 14 Based on the bioinformatics study and previous work, we attempted to verify the hypothesis that the miR-17–92 cluster may share the same expression unit. We searched for the genome location and possible co-expressed mRNA and found the six pre-miRNAs within about 1 kb on human chromosome 13; the possible co-expressed mRNA is human chromosome 13 open reading frame 25 (C13 orf25), transcript variant 2 mRNA, indicating that the miR-17–92 cluster members are intronic miRNAs. Analysis of clustered miRNA expression profiles suggested that the six clustered miRNAs may have the same promoter element. 5 15 No human miR-430 is registered online at present, but miR-17, miR-20, and miR-106a have the same sequence at nucleotides 2–8. Giraldez thought this is most important for recognition. In our study, members of the miR-17–92 cluster were highly expressed in the nervous system, but no human miRNA expression homologous to miR-430 was detected by Northern blot with a pre-miR-430 probe (data not shown). In humans, it is not miR-430 that plays a critical role in nervous system morphogenesis. So these results indicated that there are differences in the mechanisms of brain morphogenesis between human and zebra fish. 16 c 17