Identification and Characterization of Pig Embryo MicroRNAs by Solexa Sequencing

Authors


Author’s address (for correspondence): Xiaochun Tang, College of Animal Science and Veterinary Medicine, Jilin University, Changchun 130062, China, E-mail: xiaochuntang@jlu.edu.cn

Contents

MicroRNAs (miRNAs) are a class of small, non-coding RNAs of approximately 22 nucleotides in length that regulate gene expression by binding to the 3′-untranslated regions of target mRNAs. It is now clear that miRNAs are involved in many biological processes, including proliferation, differentiation and regulation of gene expression during early embryonic development. The miRBase 16.0 (2010) shows that there are 175, 673, 408 and 1048 annotated miRNAs for Caenorhabditis elegans, Mus musculus, Rattus norvegicus and Homo sapiens, respectively. However, there are only 211 miRNAs described for Sus scrofa. In particular, the full set of miRNAs and their expression patterns are still poorly understood in the embryo. Therefore, we combined Solexa sequencing with computational techniques to analyse the sequences and relative expression levels of S. scrofa miRNAs at embryonic day 33 (E33). Of the distinct miRNAs identified, 76 previously known miRNAs and 194 candidate miRNAs were identified in head, and 77 known miRNAs and 130 predicted candidate miRNAs were identified in organ region. Furthermore, we performed additional investigation for identifying the potential target mRNAs using PicTar and TargetScan. Concurrent function analysis suggested that highly expressed miRNAs are mostly involved in the development of nerves, cerebrum, muscle and organs. Our results provide useful information for the investigation into embryonic miRNAs of pig and provide a valuable resource for investigators interested in the regulation of embryonic development in pigs and other animals.

Introduction

MicroRNAs (miRNAs) are abundant, endogenous, non-coding RNAs of approximately 22 nucleotides (nt) in length (Ambros 2004; Bartel 2004; Bentwich 2005). They can bind target mRNAs and block their expression by inhibiting their translation or targeting the mRNA for degradation (Ambros 2004; Kim and Nam 2006). It is estimated that 1–5% of the genes in the genome encode for miRNAs (Lim et al. 2003), which may regulate up to 30% of all genes (Lewis et al. 2003). The first discovered miRNA was lin-4 (Lee et al. 1993), and the fact that lin-4 does not code for protein did not cause much concern. However, since the discovery of let-7 (Reinhart et al. 2000), thousands of miRNAs have been identified experimentally or computationally from a variety of species.

The pig (Sus scrofa) has tremendous biomedical importance as a model organism because it has closer phylogenic proximity to humans than mouse or other animals. Over the last few years, thousands of miRNAs from various organisms have been discovered. The majority of miRNAs are evolutionarily conserved across species and are important regulators of molecular and cellular processes, such as brain morphogenesis (Giraldez et al. 2005) and cardiomyocyte proliferation/differentiation (Zhao et al. 2005). Additional studies have shown that fat metabolism has a direct relationship with miR-14 (Xu et al. 2003), miR-143 (Esau et al. 2004) and miR-122 (Esau et al. 2006). However, little is known about the miRNA components involved in pig development.

Several research groups had applied a deep sequencing approach to successfully discover 449 new chicken miRNAs (Glazov et al. 2008) and to identify the expression levels of 212 annotated miRNAs in the porcine skeletal muscle (Nielsen et al. 2010). Similarly, a Solexa sequencing approach had been used to identify 113 amphioxus miRNAs genes (Chen et al. 2009); especially, recently researchers had successfully reported 112 conserved and 61 candidate novel porcine miRNAs from 16 different porcine tissues (Xie et al. 2011) and extended the repertoire of pig miRNAome to 867 encoding for 1004 miRNAs, of which 777 are unique (Li et al. 2010). The identification and characterization of miRNAs that are expressed at critical stages of development would provide valuable insight into the roles of miRNAs during organogenesis. Therefore, we used Solexa sequencing to systemically analyse the expression profiles of porcine embryonic miRNAs at E33 (from head and organ regions). This study will enable us to better understand the role and function of miRNAs during pig embryonic development.

Materials and Methods

Animal collection and RNA sample preparation

Our studies included two RNA samples isolated from the embryo of Large White pigs at E33, from head and organ regions. The experimental protocol used in this study was approved by the Laboratory Animal Resource Center of Jilin University and was performed in accordance with the Animal Care and Use Statute of China. After surgically removing the embryos from the sow and removal of the amnion, head and organ regions were separated using scissors. Total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA, USA). The concentration and purity of RNA samples were determined by measuring the A260/A280 ratio using NanoDrop ND1000 spectrophotometer (Thermo Scientific, Hudson, NH, USA).

Solexa sequencing experiments

Total RNA from head and organ regions isolated from three individual embryos at E33 were pooled. Approximately 40 μg of total RNA representing head or organ region was submitted to a commercial service provider (BGI, Shenzhen, China) for Solexa sequencing. The sequencing procedure was conducted as previously described (Chen et al. 2008, 2009). Briefly, after polyacrylamide gel electrophoresis (PAGE) purification of small RNA (sRNA) molecules, enrichment for molecules in the range of 18–30 nt and ligation to a pair of Solexa adaptors at their 5′ and 3′ ends, the sRNA molecules were reverse-transcribed as recommended by the manufacturer and amplified using the adaptor primers for 18 cycles, and the fragments of approximately 90 bp were isolated from agarose gels. The purified DNA was used directly for library generation and sequencing by Solexa’s proprietary sequencing-by-synthesis (SBS) method. Based on small RNA digital analysis of high-throughput sequencing, SBS can reduce regional missing that was led by secondary structure. After removal of adaptor sequences, discarding contaminated reads and low-quality reads, the high-quality and clean reads were processed for computational analysis.

Bioinformatics analysis of sequences

First, Solexa reads were aligned against the S. scrofa genome using the soap (Short Oligonucleotide Alignment Program) (Li et al. 2008) from the UCSC genome browser (http://genome.ucsc.edu/) (Kuhn et al. 2009) to analyse the distribution of the sRNAs in the genome. Next, these reads were compared with pre-miRNAs in the miRBase using blastn to identify conserved miRNAs in pig. To further analyse the RNA secondary structures of the matched Solexa reads, we used the Mfold program (http://www.bioinfo.rpi.edu/applications/mfold) to analyse the 100-nt flanking sequences from the pig genome (Hofacker 2003; Zuker 2003). Stem-loop hairpins were considered typical only when they fulfilled three criteria: mature miRNAs are present in one arm of the hairpin precursor that lack large internal loops or bulges; the secondary structure of the hairpins is steady with the free energy of hybridization lower than −20 kcal/mol; and hairpins located in intergenic region or intron (Chen et al. 2009). The genes whose sequences and structures satisfied all of these criteria were considered to be candidate miRNA genes.

Poly(A)-tailed reverse transcription PCR

Assays to verify mature miRNAs were conducted as previously described (Fu et al. 2006; Ro et al. 2006). Briefly, miRNAs were purified using the miRcute miRNA isolation kit (Tiangen, Beijing, China). The polyadenylation reaction was performed with 2 μg of miRNA and 5 U of poly(A) polymerase (NEB, Beverly, MA, USA). The reverse transcription (RT) PCR was performed using 2 μg of poly(A)-tailed miRNA and 1 μl of RT primer: 5′-ATTCTAGAGGCCGAGGCGGCCGACATGd(T)30(A,G,orC)(A,G,C,orT)-3′, 200 U of SuperScript III (Invitrogen), and the reaction mixture was incubated for 60 min at 50°C. Then, amplification of the miRNAs was performed using a miRNA-specific primer and a universal reverse primer: 5′-ATTCTAGAGGCCG AGGCGGCCGACATGT-3′ for 25 cycles. The PCR products were analysed on 12% polyacrylamide gel with ethidium bromide staining.

Northern blotting

Approximately 50 μg of total RNA was loaded onto denaturing 15% urea–polyacrylamide gels. After electrophoresis, the RNA was transferred to a HybondTM-N+ nylon membrane (Amersham, San Francisco, CA, USA). Pre-hybridization of the membranes was carried out for 60 min at 42°C using ExpressHyb Hybridization Solution (Clontech, Palo Alto, CA, USA). Then, hybridization in the presence of 3′-DIG-labelled specific DNA probes was performed overnight at 42°C. After membrane was incubated in 1% blocking solution (Roche, Mannheim, Germany) for 30 min followed by incubation in antibody solution (anti-DIG, alkaline phosphatase–conjugated antibody; Roche) for 30 min, the blots were incubated with the chemiluminescent substrate CSPD (Roche) and exposed to Kodak BioMax MR Film.

Quantitative real-time PCR

Real-time PCR was carried out using a BioEasy SYBR Green I Real-time PCR Kit (Bioer, Hangzhou, China). First-strand cDNA synthesis of miRNA was performed by poly(A)tailed RT-PCR. The mixture was incubated at 94°C for 2 min, followed by 40 cycles of 94°C for 10 s, 60°C for 15 s and 72°C for 30 s with Mx3000P (Stratagene, La Jolla, CA, USA). The expression levels of miRNAs were measured by Ct (threshold cycle). The relative expression level was generated using the equation inline image. The U6 gene was used as an endogenous control. U6_F: 5′-GCTTCGGCAGCACATATACTAAAAT-3′ and U6_R: 5′-CGCTTCACGAATTTGCGTGTCA T-3′.

Statistical analysis

All values are reported as means ± SD. The data regarding real-time PCR were analysed by one-way anova with spss 16.0 (SPSS Inc, Chicago, IL, USA), and p < 0.05 was considered to be statistically significantly different.

Results

Construction of a small RNA library by Solexa sequencing and bioinformatics analysis

To determine the identity and abundance of miRNAs from a specific stage of pig embryonic development, an sRNA library was sequenced using Solexa technology at E33. The sequence reads were clustered into unique sequences, of which the vast majority were 22 nt in length. After removal of the adaptor sequences, and contaminated and low-quality reads, head and organ region libraries contained 10 998 408 and 10 813 199 filtered clean reads, respectively (Table 1). Next, all Solexa reads were aligned with the S. scrofa genome. The results indicated that 9 065 331 (541 582 unique sRNAs) and 8 692 336 (412 150 unique sRNAs) reads were perfectly matched to the pig genome in head and organ regions, respectively. These reads were compared with the pre-miRNAs in miRBase to obtain information about annotated miRNAs. Selection of the sRNAs in NCBI annotated the Solexa reads and excluded the non-miRNAs as much as possible (Table 2). After having scanned the S. scrofa genome for hairpin structures using Mfold, we identified 76 mature miRNAs (Table 3) and 194 candidate miRNAs in head region (Table S1). In organ region, 77 mature miRNAs (Table 4) and 130 candidate miRNAs (Table S2) were identified. Sequencing-based miRNA expression profiling has often been used as a tool to measure the relative abundance of miRNAs (Landgraf et al. 2007). The sequencing frequency of the four most abundantly expressed miRNAs in head (ssc-miR-9, ssc-let-7c, ssc-miR-103 and ssc-miR-30a) constituted 75.55% of the total miRNA reads from this region. The most abundantly expressed miRNAs from organ region (ssc-miR-140*, ssc-miR-1, ssc-let-7, ssc-miR-21, ssc-miR-30a) constituted 61.72% of the total miRNA reads from this region, suggesting that they might be very important in pig embryonic development. In contrast, the sequencing frequencies of ssc-miR-196, ssc-miR-224 and ssc-miR-326 were extremely low in head library, and ssc-miR-105, ssc-miR-325 and ssc-miR-301 were low in organ library. It is possible that the miRNAs were expressed at very low levels in limited cell types, and/or under limited circumstances at this specific stage, consistent with previous studies in porcine adipose tissue (Li et al. 2011). In addition, of the identified sequences, let-7 was abundantly expressed, which is consistent with let-7 being ubiquitously expressed in the tissues (Huang et al. 2008) and with the requirement of ssc-let-7 for the timing of cell fate determination (Pasquinelli et al. 2000; Reinhart et al. 2000).

Table 1.   Number of sequencing reads from the head and organ regions at E33 by small RNA sequencing
 The head regionThe organ region
Total reads11 617 07911 458 732
High-quality reads11 400 04011 256 616
Clean reads10 998 40810 813 199
Table 2.   Summary of small RNAs that matched with corresponding GenBank non-coding RNAs
 Total readsaUnique sRNA geneaTotal readsbUnique sRNA geneb
  1. aThe RNA sample from the head region.

  2. bThe RNA sample from the organ region.

rRNA470 39019039599 45122041
snRNA114 977737577 5636537
snoRNA60 748612548 7995820
tRNA334 16512853550 71416489
Other10 018 1288042719 536 672626049
Total10 998 40884966310 813 199676936
Table 3.   Identity and abundance of porcine miRNAs from head tissue at E33
MicroRNACountMicroRNACountMicroRNACountMicroRNACount
ssc-mir-9-11 504 449ssc-mir-101a-235 060ssc-mir-2173909ssc-mir-204833
ssc-mir-9-21 504 445ssc-mir-1729 685ssc-mir-106a3674ssc-mir-210829
ssc-let-7c741 019ssc-mir-18327 895ssc-mir-30b3606ssc-mir-214652
ssc-mir-103367 926ssc-mir-26a21 488ssc-mir-1853197ssc-mir-122450
ssc-mir-30a210 794ssc-mir-2418 115ssc-mir-27a3026ssc-mir-139350
ssc-mir-140*188 896ssc-mir-130a17 838ssc-mir-1532964ssc-mir-23a338
ssc-mir-99b131 479ssc-mir-199b17 516ssc-mir-2212918ssc-mir-145264
ssc-let-7f117 228ssc-mir-16-217 026ssc-mir-181c2861ssc-mir-325212
ssc-mir-18487 211ssc-mir-16-117 005ssc-mir-2162735ssc-mir-215149
ssc-mir-786 110ssc-mir-18616 734ssc-mir-1072596ssc-mir-205139
ssc-mir-181a84 475ssc-mir-1a14 813ssc-mir-105-12498ssc-mir-105-2130
ssc-mir-2158 790ssc-mir-14012 932ssc-mir-15b2022ssc-mir-133a87
ssc-mir-148a56 065ssc-mir-30c11 655ssc-mir-951732ssc-mir-29c65
ssc-mir-125b55 677ssc-mir-124a-19870ssc-mir-321713ssc-mir-29b62
ssc-mir-135-137 908ssc-mir-124a-29869ssc-mir-3231500ssc-mir-45061
ssc-mir-135-237 886ssc-mir-1288182ssc-mir-19a1173ssc-mir-34a44
ssc-mir-2036 921ssc-let-7i6276ssc-mir-1361052ssc-mir-19640
ssc-mir-101a-136 189ssc-mir-184857ssc-mir-146b877ssc-mir-22414
ssc-mir-181b35 102ssc-mir-15a4237ssc-mir-503871ssc-mir-3268
Table 4.   Identity and abundance of porcine miRNAs from organ tissue at E33
MicroRNACountMicroRNACountMicroRNACount
ssc-miR-1a613 588ssc-miR-169700ssc-miR-181c714
ssc-miR-140*320 618ssc-miR-169695ssc-miR-221678
ssc-let-7c216 451ssc-miR-130a8248ssc-miR-124a532
ssc-miR-30a212 657ssc-miR-146b8169ssc-miR-124a532
ssc-miR-21142 059ssc-miR-27a6809ssc-miR-32483
ssc-miR-148a138 171ssc-miR-76255ssc-miR-136481
ssc-miR-99b101 775ssc-let-7i5164ssc-miR-19a448
ssc-miR-10397 903ssc-miR-30b4111ssc-miR-139387
ssc-let-7f63 969ssc-miR-133a3677ssc-miR-323386
ssc-miR-2462 700ssc-miR-1453605ssc-miR-217268
ssc-miR-199b*55 791ssc-miR-5033396ssc-miR-95216
ssc-miR-9-141 925ssc-miR-2142943ssc-miR-450181
ssc-miR-9-241 925ssc-miR-1842730ssc-miR-105-1180
ssc-miR-18339 742ssc-miR-15a2543ssc-miR-34a178
ssc-miR-181a25 909ssc-miR-182533ssc-miR-216119
ssc-miR-26a22 583ssc-miR-1282458ssc-miR-20595
ssc-miR-101a18 108ssc-miR-2152252ssc-miR-29b40
ssc-miR-101a17 568ssc-miR-1852038ssc-miR-15335
ssc-miR-2016 298ssc-miR-2101434ssc-miR-29c31
ssc-miR-14014 851ssc-miR-106a1392ssc-miR-20430
ssc-miR-125b13 672ssc-miR-1961244ssc-miR-22430
ssc-miR-1713 631ssc-miR-15b1066ssc-miR-32617
ssc-miR-12213 470ssc-miR-107996ssc-miR-105-212
ssc-miR-181b12 636ssc-miR-23a987ssc-miR-3254
ssc-miR-18610 153ssc-miR-135762ssc-miR-3012
ssc-miR-30c9832ssc-miR-135760  

Detection of embryonic miRNA expression with poly(A)-tailed RT-PCR and Northern blotting

After sequencing, we carried out poly(A)-tailed RT-PCR to verify the relative abundance of miRNAs. We selected these miRNAs for further analysis, which were highly expressed in head or organ region and potentially related to embryonic development. With the exception of let-7, from head library, we chose to assay ssc-miR-9, ssc-miR-103, ssc-miR-140*, ssc-miR-184 and ssc-miR-30a (Fig. 1a) and the candidate miRNAs ptztb-m0026, ptztb-m0059, ptztb-m0190, ptztb-m0146 and ptztb-m0191 (Fig. 1b). From organ library, we chose to verify ssc-miR-1, ssc-miR-24, ssc-miR-183, ssc-miR-26a and ssc-miR-101a (Fig. 1c) and the candidate miRNAs ptznz-m0117, ptznz-m0087, ptznz-m0065, ptznz-m0110 and ptznz-m0005 (Fig. 1d). In agreement with our sequencing data, we showed that these miRNAs were indeed clearly expressed in pig embryo. We found that the outcomes for annotated miRNAs were better than those of candidate miRNAs likely because the expression levels of annotated miRNAs were higher than those of candidate miRNAs. In addition, we found that the highly expressed miRNAs from organ region were also expressed in head region (Fig. 1e,f) and that the highly expressed miRNAs from head region were also expressed in organ region (Fig. 1g,h). These results, which were also consistent with our sequencing data, suggest two possibilities: these miRNAs are expressed in both regions, such as in nervous system of the cerebrum or in an organ of the body; and the miRNAs expressed in head region have obvious effects in this region, but could have little effect in organ region and the vice versa. Further experiments are needed to investigate these possibilities.

Figure 1.

 Verification of the highly expressed miRNAs from head and organ regions with poly(A)-tailed RT-PCR. The lengths of the products were approximately 80 bp. (a, b) 10 miRNAs selected from head region were verified in head region. (c, d) 10 miRNAs selected from organ region were verified in organ region. (e, f) 10 miRNAs selected from organ region were verified in head region. (g, h) 10 miRNAs selected from head region were verified in organ region. ladder: 50 bp

As an additional method to confirm the outcomes of the miRNA deep sequencing, northern blotting was used. Several groups have established a non-isotopic northern blotting analysis method for miRNA detection using 3′-DIG-labelled RNA oligo probes (Ramkissoon et al. 2006). Because ssc-miR-103, ssc-miR-140* and ssc-miR-1 were highly expressed by sequence analysis, we applied northern blotting to validate the sequencing data of ssc-miR-1 from organ region (Fig. 2a) and of ssc-miR-103 and ssc-miR-140* from head region (Fig. 2b,c). The results confirmed that miR-103, miR-140* and miR-1 are abundant in embryonic tissues.

Figure 2.

 Northern blots showing the detection of miR-103 and miR-140* in head region and miR-1 in organ region. 3′-DIG-labelled DNA oligonucleotide probes detected miR-1 (a), miR-103 (b) and miR-140* (c)

Additionally, for highly expressed miRNAs in embryo, we determined the relative amount of miRNAs by quantitative real-time PCR to confirm the expression of miRNAs determined by Solexa sequencing. We compared the highly expressed annotated miRNAs and candidate miRNAs in head region (Fig. 3a,b) and in organ region (Fig. 3c,d). There was a consistency in results between the quantitative real-time PCR assays and Solexa sequence for the 20 miRNAs analysed. We also compared the relative miRNA expression levels in adult tissues (heart, liver, spleen, lung, kidney, muscle and brain) for miR-1 (Fig. 3e), miR-103 (Fig. 3f) and miR-140* (Fig. 3g). The results showed that the expression of miR-1, miR-103 and miR-140* were highest in muscle, brain and lung, respectively. Thus, these miRNAs may also have important roles in these tissues of the adult pigs.

Figure 3.

 Relative quantification of highly expressed miRNAs in head and organ regions. (a, b) The relative expression of 10 annotated miRNAs and five candidate miRNAs in head region. (c, d) The relative expression of 10 annotated miRNAs and five candidate miRNAs in organ region. Expression profiles of ssc-miR-1 (e), ssc-mir-103 (f) and ssc-miR-140* (g) in various tissues of adult Large White pigs. The miRNA expression levels were normalized to U6 (n = 3, mean ± SD). *Values within groups are significantly different (p < 0.05)

Secondary structure prediction of the precursor sequences for candidate porcine miRNAs

One of the important criteria that distinguish miRNAs from other endogenous sRNAs is the ability of the flanking sequences to adopt a hairpin precursor structure that is excised during processing by Dicer (Ambros et al. 2003). The sequences that we identified as candidate miRNAs did not exist in miRBase and are potential, novel miRNAs. To annotate these newly identified porcine miRNAs, sequence reads that matched genomic sequences were selected, and the 40 nt that flanked each side was extracted (Zuker 2003). The predicted stem-loop structures of the 10 most abundant candidate miRNAs are shown in Table 5.

Table 5.   Sequences and predicted stem-loop structures for 10 candidate miRNAs
NameSequences and structures
ptztb-m0026CTGTTCTCTTGAGAGATCAGAGGCGCAGAGTGCGTCAATGTCAGTGAAGCCTGTGCCTTTTACCTCTTTAAGAGCGCACGG
.(((.((((((((((...((((((((((.((...(((.......))).))))))))))))...)))).)))))).)))... (−31.30 kcal/mol)
ptztb-m0059GCCCTGGCCCCAGGGTCCTCTTGTCTGAGCCCCAGAAAGAGGAGAGAGCGCTGGGTCACAGAAGAGGGTCTGGGGGACAGGTT
..((((.((((.((..(((((..((((.(.(((((...(.........).))))).).)))))))))..)).)))).)))).. (−40.10 kcal/mol)
ptztb-m0190TGAAAAGTTCCGTCAACCATCCAGCTGTTTGGGGTGATGCAAACAAACATCTGGTTGGTTGAGAGAATTTTTTACT ((((((((((..(((((((.((((.((((((..(.....)...)))))).)))).)))))))..)))))))))).. (−32.20 kcal/mol)
ptztb-m0146GCTGTCCCCATGGCACAGGGTCCAGCTGTCGGCTGTAATACCCGATGGGTCGATGATGGTCCCTGTGTTTGGGGCGAGCAC (((..(((((.(((((((((.(((..((((((((.............)))))))).))).))))))))))))))..))).. (−41.72 kcal/mol)
ptztb-m0191TCTCTTGTGTTAAGGTGCATCTAGTGCAGTTAGTGAAGCAGCTTAGAATCTACTGCCCTAAATGCCCCTTCTGGCACAGGCTGCC...((((((((((((.((((.(((.(((((..((.(((...)))...))..))))).))).)))).)))..)))))))))..... (−27.60 kcal/mol)
ptznz-m0117CACAGCTTTCCAGAGGTTTCCAGCTTATAGGATATTCATATGGTAGTTCTGTGAACTAGAAACCTCTGGAAAGTGGTGT
(((.((((((((((((((((.((.(((((((((..((....))..))))))))).)).)))))))))))))))).))). (−40.20 kcal/mol)
ptznz-m0087TGTGGTGCCTCCATTCCTTCGTCTGTGCACTAGAAAAATAACAGTACCTAGGGCACAGGATGGGATGAGGAGTAAACAGGG (((..((((..((((((....(((((((.((((..............)))).)))))))..))))))..).))).)))... (−27.84 kcal/mol)
ptznz-m0065CTAGGCCACGGGGGAACTCCCAGACCAGCTTCCACTCCCTTCAAGAAGCTGGTCTGGGAGTTCCCGGGTGGCCTAGTA
(((((((((..(((((((((((((((((((((............)))))))))))))))))))))..))))))))).. (−62.20 kcal/mol)
ptznz-m0110AGACATCGCTTCTGGTCCAGACACTGTGGAGCTGAAGAGACCAACACTTCTCAGTGTTTGAACCAGAAGCTCTGTCTGCA
(((((..(((((((((.(((((((((.((((.((..(....)..)))))).))))))))).)))))))))..)))))... (−40.40 kcal/mol)
ptznz-m0005GGTTCCGTTTACTTTCCCGGGATTTGGAGCGCTTCCGCACGCGCTCCTTGTCCCGAGACCGCGACCGCGACCG
((((.((.(..(..((.((((((..(((((((........)))))))..)))))).))..)..).)).)))). (−35.20 kcal/mol)

miRNA target predictions, developmental function analysis and Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis

To begin to elucidate the cellular functions of the 10 most abundant, annotated miRNAs in embryo, miRNA target prediction was performed by searching for the presence of conserved 7-mer sites in the 3′-UTR of mRNAs. In animal genomes, miRNA targets are difficult to predict because the targets of miRNAs generally display only partial complementarity to the mature miRNA sequence (Millar and Waterhouse 2005; Carthew 2006). Therefore, in an attempt to reveal their functions in pig embryonic development, the output results of the algorithms PicTar (Krek et al. 2005) and TargetScan (Lewis et al. 2003) were subsequently compared to increase the specificity of the predictions. The genes were chosen according to their functional annotations to identify pathways that are actively regulated by miRNAs during embryonic development. We chose 60 target genes related to embryonic development (Table 6). The results showed that many functions in the pig involve several miRNA family groups. These functions include neuronal development, embryonic limb morphogenesis, skeletal development, determination of left/right symmetry and embryonic anterior compartment specification. Meanwhile, Because KEGG (Kyoto Encyclopedia of Genes and Genomes) is a systematic network database to analyze complex gene functions and cellular pathways, especially large-scale biological interpretaion in genomics, transcriptomics, proteomics, the target genes were classified according to KEGG functional annotations to identify pathways that were actively regulated by miRNA in porcine embryo. A total of 20 possible pathways were revealed (Table 7). It appeared that the enriched miRNAs in porcine embryo were intensively involved in not only the development of the nervous system such as neurotrophin-signalling and cholinergic synapse but also the development of the organ such as long-term potentiation and osteoclast differentiation. Their roles involved regulations of important signalling pathways including MAPK, TGF-β, pathways in cancer and Wnt signalling pathway, as well as regulations of intercellular activities such as focal adhesion, cytokine-cytokine receptor interaction and gap junction.

Table 6.   Potential targets related to embryonic development for the partial identified miRNAs in porcine embryo and function analysis
miRNATarget gene related to embryonic developmentFunctions
ssc-miR-9Fibrillin-1 precursor (FBN1),
Homeobox protein engrailed-2 (Hu-En-2),
Protein patched homolog 1 (PTC1),
Homeobox protein Hox-A11 (HOXA11),
Dual specificity testis-specific protein kinase 2 (TESK2),
platelet – derived growth factor C precursor (PDGFC)
Heart development, skeletal development, spermatogenesis, embryonic limb morphogenesis, hindbrain development
ssc-miR-103SNF-related serine/threonine-protein kinase (SNRK),
Axis inhibition protein 2 (Axin-2),
Delta-like protein 1 precursor (DLL1),
Dual-specificity mitogen-activated protein kinase kinase 1 (MAP2K1),
Pre-B-cell leukemia transcription factor 3 (PBX3),
Cyclin-dependent kinase 5 activator 1 precursor (CDK5R1)
Neuron differentiation, brain development, myeloid cell differentiation, left/right symmetry determination, osteoblast differentiation
ssc-miR-140*Jagged-1 precursor (JAG1), B-cell lymphoma 9 protein (Bcl-9),
TGF-B superfamily receptor type I (TSR-I),
Early growth response protein 2 (EGR2),
Ubiquitin-protein ligase E3A (UBE3A),
Histone deacetylase 4 (HDAC4)
Organ morphogenesis
peripheral, nervous system development, neural crest cell migration, striated muscle development
ssc-miR-184Frizzled-1 precursor (FZD1),
Band 4.1-like protein 5 (EPB41L5),
Immunoglobulin superfamily member 8 precursor (IGSF8),
Brain-specific angiogenesis inhibitor 2 precursor (BAI2),
DNA topoisomerase 2-beta (TOP2β),
RNA-binding protein Musashi homolog 1 (MSI1)
Paraxial mesoderm development, forebrain development, striated muscle development, axial mesoderm development, nervous system development
ssc-miR-30aSerine/threonine-protein kinase receptor R3 precursor (ACVRL1),
SOX11-Transcription factor (SOX-11),
Adenomatous polyposis coli protein (APC),
Pituitary homeobox 2 (PITX2),
Amyloid beta A4 precursor protein-binding family A member 1 (APBA1),
Homeobox protein DLX-6 (DLX6)
Anterior/posterior pattern formation, dorsal/ventral pattern formation, embryonic limb morphogenesis, cell proliferation
ssc-miR-1Homeobox protein Hox-B4 (HOXB4),
Neurogenic locus notch homolog protein 3 precursor (NOTCH3),
Ephrin-B2 precursor (EFNB2),
Endothelin-1 precursor (EDN1),
Mab-21-like protein 1 (MAB21L1),
Ras GTPase-activating protein 1 (RASA1)
Forebrain development, dorsal/ventral pattern formation, neuron migration, neuron apoptosis
ssc-miR-24Semaphorin-4G precursor (SEMA4G),
Hepatocyte nuclear factor 1-beta (HNF-1β),
Chordin precursor (CHRD),
Myogenin,
Neurogenic differentiation factor 1 (NEUROD1),
Delta-like protein 1 precursor (DLL1)
Anterior/posterior pattern formation, cell differentiation/striated muscle development, dorsal/ventral pattern formation, hemopoiesis, somite specification
ssc-miR-183Low-density lipoprotein receptor–related protein 6 precursor (LRP6),
Insulin receptor substrate 1 (IRS1),
Serine/threonine-protein phosphatase 2A catalytic subunit alpha isoform (PPP2CA),
Zinc finger protein 3 (ZIC3)
Protein kinase C alpha type (PRKCA),
Muscleblind-like protein (MBNL1)
Gastrulation, anterior/posterior pattern formation, left/right symmetry determination, mesoderm development
ssc-miR-26aJagged-1 precursor (JAG1),
B-cell translocation gene 1 protein (BTG1),
Inhibin beta B chain precursor (INHBB)
Phosphatidylinositol-3,4,5-trisphosphate 3-phosphatase and dual-specificity protein phosphatase (PTEN), E1A-associated protein p300 (EP300), Phosphatidylinositol 3-kinase regulatory subunit gamma (PIK3R3)
Organ morphogenesis, central nervous system development, polarity of embryonic epithelium establishment, skeletal development, myoblast/endothelial cell differentiation
ssc-miR-101Cyclin-dependent kinase 5 activator 1 precursor (CDK5R1),
Fibrillin-2 (FBN2),
Neurogenic differentiation factor 1 (NEUROD1),
Zinc finger homeobox protein 1b (ZFHX1B),
Nemo-like kinase (NLK),
Transcription factor SOX-9 (SOX9)
Neuron differentiation, dorsal/ventral pattern formation, axonal fasciculation, heart development
Table 7.   KEGG pathways probably regulated by top 10 most abundant miRNAs in porcine embryo
TermCountp-value
  1. Count represents targeted genes involved in the term.

hsa04010: MAPK signalling pathway211.50E-10
hsa04350: TGF-β signalling pathway193.18E-08
hsa04060: Cytokine–cytokine receptor interaction165.67E-08
hsa04540: Gap junction161.12E-07
hsa04380: Osteoclast differentiation144.50E-06
hsa04110: Cell cycle142.50E-06
hsa05200: Pathways in cancer271.40E-05
hsa05202: Transcriptional misregulation in cancer152.70E-05
hsa04725: Cholinergic synapse152.10E-05
hsa04510: Focal adhesion163.90E-05
hsa04810: Regulation of actin cytoskeleton165.50E-05
hsa05215: Prostate cancer175.71E-05
hsa05218: Melanoma185.52E-04
hsa05214: Glioma172.97E-04
hsa04310: Wnt signalling pathway183.99E-04
hsa04330: Notch signalling pathway161.30E-04
hsa04722: Neurotrophin signalling pathway141.29E-03
hsa04270: Vascular smooth muscle contraction131.57E-03
hsa04660: T cell receptor signalling pathway132.04E-03
hsa04720: Long-term potentiation143.14E-03

These analyses illustrate some of the possible roles of the highly expressed miRNAs in signalling pathways during embryonic development. The discovery of miRNAs in pig is important for the identification of miRNA-dependent regulatory networks that are essential for developmental processes and physiology in this species.

Discussion

miRNAs represent a new class of factors that can regulate gene expression. To date, miRNA expression had not been analysed in the pig embryo, and the roles of miRNAs in embryonic development remained unclear; however, research in embryonic development is an important issue in life science. The study of the embryos of various animals will enable us to more fully understand the mechanisms of embryonic development through identification of novel miRNAs, their targets and their various functions. Several groups have shown that miR-206 targets the p180 subunit of DNA polyα to negatively regulate its translation and to promote muscle differentiation (Kim et al. 2006). In Drosophila, miR-1 influences the Notch signalling pathway through inhibition of the translation of Delta, a ligand of Notch, to promote the differentiation of cardiac cells (Kwon et al. 2005). In humans, miR-9 regulates proliferation and migration of human neural progenitor cells (Delaloy et al. 2010; Uchida 2010).

Many technologies have been developed for miRNA profiling, including quantitative real-time PCR, microarray analyses. These methods have been used successfully in a variety of studies, but they still have some technical limitations, apart from time-consuming and labour-intensive. Recently, it is reported that 227 conserved miRNAs (of which 59 were novel) and 66 potential porcine miRNAs were successfully identified in porcine adipose tissue with Solexa sequencing (Li et al. 2011). For this reason, we also used miRNA Solexa sequencing to analyse head region and organ region samples that we collected at E33 (a stage at which morphogenesis has been basically completed) from Large White pigs. Our Solexa sequencing approach for the discovery of miRNAs that are expressed at E33 of pig confirmed the expression of 76 known miRNAs and 194 predicted miRNAs in head region, and 77 known miRNAs and 130 predicted miRNAs in organ region. We used quantitative real-time PCR to verify the expression of 10 annotated miRNAs and 10 candidate miRNAs. The results were consistent with our sequencing results, which indicates that Solexa sequencing does allow for the successful discovery of miRNAs from pig with high accuracy and efficiency. The higher levels of miRNA expression in embryonic tissues imply that these miRNAs play critical roles in embryonic development.

In this study, the predicted potential target genes for the miRNAs related to pathways involved in embryonic development indicated that these miRNAs likely have a role in the regulation of embryonic development. For example, miR-103 was predicted to target the gene AXIN2, which can combine with adenomatous polyposis coli (APC) protein to stabilize beta-catenin to regulate the transcriptional activity of downstream genes, such as c-myc, MMP7, ID2, CD44, EphB2 and FGF20. miR-1 can target neurogenic locus notch homolog protein 3 precursor (Notch3) to negatively regulate neuronal differentiation, forebrain development and other processes. These observations suggest some possible mechanisms by which these miRNAs could participate in the regulation of pig embryo development, although this remains to be validated experimentally. From our function analysis, we found that EP300 (E1A-associated protein p300) is a target of ssc-miR-30a and miR-206, which can regulate lung development. At the same time, from the pathway analysis, the most over-represented miRNA targets belonged to the MAPK and TGF-β signalling pathway, which is known to be involved in widely cellular functions such as proliferation, apoptosis, differentiation. Furthermore, pathways associated with the regulation of focal adhesion, cell cycle gap junctions indicated the role of the highly expressed miRNAs in the regulation of cell motility, cell proliferation and the cytoskeleton. Moreover, the neurotrophin signalling pathway and long-term potentiation suggested that the top 10 miRNAs participate in nervous system development and function. These pathway and biological process analyses illustrate some of the possible roles of the highly expressed miRNAs in porcine embryos.

In summary, we identified porcine miRNAs that are expressed during development (E33) using Solexa sequencing technology. Our findings provide a deeper understanding of this accurate and efficient approach for miRNA discovery and will aid future identification of miRNAs in other species. More importantly, our study represents the first attempt to conduct a deep analysis of miRNAs in pig embryo at a specific development stage and provides the basis for future analysis of miRNA function during embryonic development. The miRNA expression signatures are clearly distinct in embryonic development, indicating their role in organogenesis of the pig embryo.

Acknowledgements

We thank everyone involved for their assistance with this project. This study was supported by a grant from the National Key Basic Research Program during the eleventh 5-year plan period (2009CB941001).

Conflict of interest

None of the authors have any conflict of interest to declare.

Author contributions

Zhou Yan performed real-time PCR and northern blotting, provided all figures and wrote the manuscript; Tang Xiaochun, Song Qi, Wang Huan, Wang Haijun and Ouyang Hongsheng analysed and interpreted the data; and Prof. Pang and Jiao were in charge of study design and supervision.

Ancillary