Genome-wide identification and analysis of miRNA-related single nucleotide polymorphisms (SNPs) in rice

Abstract

Background: MiRNAs are key regulators in the miRNA-mediated regulatory networks. Single nucleotide polymorphisms (SNPs) that occur at miRNA-related regions may cause serious phenotype changes. To gain new insights into the evolution of miRNAs after SNP variation, we performed a genome-wide scan of miRNA-related SNPs, and analyzed their effects on the stability of miRNAs structure and the alteration of target spectrum in rice.

Results: We find that the SNP density in pre-miRNAs is significantly higher than that in the flanking regions, owing to the rapid evolution of a large number of species-specific miRNAs in rice. In contrast, it is obvious that deeply conserved miRNAs are under strong purifying selection during evolution. In most cases, the SNPs in stem regions may result in the miRNA hairpin structures changing from stable to unstable status; And SNPs in mature miRNAs have great potential to have either newly created or disrupted the miRNA-target interactions. However, the total number of gained targets is over 2.5 times greater than that are lost after mutation. Notably, 12 putative domestication-related miRNAs have been identified, where the SNP density is significantly lower.

Conclusions: The present study provides the first outline of SNP variations occurred in rice pre-miRNAs at the whole genome-wide level. These analyses may deepen our understanding on the effects of SNPs on the evolution of miRNAs in the rice genome.

Figure 1

Frequency distributions of different SNP numbers (a) and SNP density (b) in pre-miRNAs.

Figure 2

Single nucleotide polymorphism (SNP) density of all rice pre-miRNAs (a), and deeply conserved miRNAs (b), and their flanking regions. The up or down flank region represents a sequence region that is equal to the length of corresponding pre-miRNA, and located immediately adjacent to the pre-miRNA. The differences of SNP density between pre-miRNA and flanking regions were assessed using the ANOVA analysis. Data are reported as the average SNP density value ± s.e. The symbols * and ** designate the significant difference of SNP density in other regions as compared to that in pre-miRNAs at the 0.05 and 0.01 level, respectively.

Figure 3

SNP density for each site in rice mature miRNA regions. Data are reported as the average SNP density value ± s.e. The different letters (a , b , c, and d) designate the significant difference of SNP density between different sites at the 0.05 level.

Figure 4

Illustration of substitution types for SNPs that occur in rice pre-miRNAs. Arrows indicate the direction of single nucleotide mutation. The ancestral state of each SNP site was derived from the Rice Haplotype Map Project Database (Huang et al. 2012b).

Figure 5

Illustration of the predicted secondary structures for SNP- and wild-type osa-miR399j (a), osa-miR394 (b), and osa-miR399b (c). The mature miRNAs are shaded in pink. The circle marks the SNP position. The three miRNAs are shown as typical examples for displaying the increased (ΔΔG = −6.90 kcal/mol), unchangeable (ΔΔG = 0.00 kcal/mol), and decreased stability of hairpin structures (ΔΔG = 6.56 kcal/mol) by SNP variations.

Figure 6

Histogram displaying the distribution of energy change (ΔΔG, kcal/mol) of miRNA hairpin structures caused by SNPs in rice pre-miRNAs.

Figure 7

Target spectrum alteration analyses of miRNAs by SNPs in their mature sequence regions. The number of targets that were lost (a) or gained (b) was presented by comparison of wild-type miRNAs with SNP-miRNAs. To facilitate the comparison, the Y-axis is shown in log (10) scale.

Figure 8

Comparison of SNP density in the precursor sequences (pre-miRNAs) and flanking regions for putative domestication-related miRNAs (DR-miRNAs) and other miRNAs in cultivated and wild rice. For comparison, 2,000 independent permutations were performed by randomly choosing 12 miRNAs from the cultivars and wild rice miRNA datasets, respectively. Data are reported as the average SNP density value ± s.e. ASR-C and ASR-W mean DR-miRNAs in cultivated and wild rice, respectively; S-C and S-W designate other miRNAs (12 miRNAs per permutation) randomly extracted from cultivars and wild rice, respectively.