Postcards from the Rice Genome: Massive Analysis of Small RNAs in Response to Environmental Stress

MicroRNAs (miRNAs) have critical functions in plant growth and development and are increasingly recognized as having important roles in plant responses to environmental stress. However, it is a considerable challenge to identify functional miRNAs and distinguish them from the large number of other short interfering RNAs, which differ in their biogenesis, mechanism of action, and function. In plants, miRNAs function as posttranscriptional regulators usually by directing cleavage of target gene mRNA transcripts, whereas short interfering RNAs typically mediate DNA methylation and chromatin remodeling. Jeong et al. (pages 4185–4207) present a large-scale analysis of small RNAs in rice (Oryza sativa) based on deep sequencing of small RNA libraries from different rice tissues and from plants grown under a variety of abiotic stress and nutrient deprivation conditions. This is combined with high-resolution RNA detection methods for characterizing miRNAs, making fine distinctions between closely related miRNA family members, and analyzing their target genes and mRNA cleavage properties. The work provides high-confidence annotation of a large collection of miRNA genes in rice, including the identification of nearly 100 new miRNAs. Moreover, it not only profiles miRNA expression in different tissues and organs and under a variety of environmental stress conditions, but experimentally validates and extends knowledge about individual miRNAs and their targets. The results illustrate how large-scale deep sequencing data sets can be followed up by lab experiments to elucidate critical information on miRNA mechanism of action and function in plant development and response to stress. More than 60 small RNA libraries were deeply sequenced, mapped to the rice genome, and analyzed through a variety of computational and experimental approaches. Methods employed included a novel plotting approach to verify miRNAs based on their abundance and distribution within stem-loop precursors, the development of a computational pipeline to identify new miRNAs, and computational target prediction. A number of miRNAs were identified that were induced or repressed in particular tissues and in response to particular stresses or nutrient deprivation, and in some cases it was shown that the expression of target genes was negatively correlated with the cognate miRNA expression pattern, as expected for a regulatory interaction. The authors show how functional specificity of distinct miRNA family members is achieved by their diversity both in expression and target selection. Differential expression patterns of distinct miRNA family members were validated by combining deep sequencing with splinted ligation–mediated detection of individual miRNAs. In addition, differential target cleavages guided by miRNA family members were analyzed using RNA ligase–mediated 5′ rapid amplification of cDNA ends (RACE) experiments. For example, it was found that miR529-5p is preferentially expressed in the panicle and guides cleavage of a number of Squamosa Promoter binding protein-Like (SPL) mRNAs, including SPL14 (see Figure). High SPL14 expression has been found to promote panicle branching linked to high yield potential in rice (Jiao et al., 2010; Miura et al., 2010), and the work of Jeong et al. suggests that SPL14 regulation by miR529a-5p might be a critical component of this effect. Interestingly, the miR529 family is conserved among grasses and thus could play a role in the regulation of plant architecture and yield in other agronomically important grasses. Differential cleavage of SPL14 by differentially expressed miR156/529 family members. Predicted and detected cleavage sites of miR156a-j and miR529a-5p are indicated by arrows and bold 5′ nucleotides on the SPL14 mRNA. Target cleavage sites were validated by RNA ligase–mediated 5′RACE. The number of cloned RACE products sequenced is shown in the histogram and indicates the prominent cleavage guided by miR529 in the panicle. (Reprinted from Figure 9 of Jeong et al. [2011].)