target genes. The expression of PHABULOSA (PHB), PHAVOLUTA (PHV) and CORONA (CNA) is significantly reduced in a jba-1D background, while REVOLUTA (REV) expression is elevated and ATHB8 is unchanged. In addition, we show that miR166 has a dynamic expression pattern in wild-type and jba-1D embryos. Our analysis demonstrates an indirect role for miRNAs in controlling meristem formation via regulation of WUS expression, and reveals complex regulation of the class III AtHD-ZIP gene family.

Two classes of small RNAs, microRNAs and short-interfering RNA (siRNAs), have been extensively studied in plants and animals. In Arabidopsis, the capacity to uncover previously uncharacterized small RNAs by means of conventional strategies seems to be reaching its limits. To discover new plant small RNAs, we developed a protocol to mine an Arabidopsis nonannotated, noncoding EST database. Using this approach, we identified an endogenous small RNA, trans-acting short-interfering RNA-auxin response factor (tasiR-ARF), that shares a 21-and 22-nt region of sequence similarity with members of the ARF gene family. tasiR-ARF has characteristics of both short-interfering RNA and microRNA, recently defined as tasiRNA. Accumulation of trans-acting siRNA depends on DICER-LIKE1 and RNA-DEPENDENT RNA POLYMERASE6 but not RNA-DEPENDENT RNA POLYMERASE2. We demonstrate that tasiR-ARF targets three ARF genes, ARF2, ARF3͞ETT, and ARF4, and that both the tasiR-ARF precursor and its target genes are evolutionarily conserved. The identification of tasiRNA-ARF as a low-abundance, previously uncharacterized small RNA species proves our method to be a useful tool to uncover additional small regulatory RNAs.auxin response factor ͉ microRNA E ndogenous noncoding small RNAs 20-25 nt in length are important regulators of gene expression in both plants and animals (1-4). Recently, two major classes of small RNAs, microRNAs (miRNAs) and short-interfering RNAs (siRNAs), have been extensively studied. These two classes of small RNAs are generated from different types of precursor molecules, are processed through distinct biochemical pathways, and function in different biological contexts. miRNAs are processed from precursor molecules that are capable of forming double-stranded RNA (dsRNA) by intramolecular pairing (5) and act as regulatory factors during growth and development (1, 6, 7). In contrast, siRNAs are processed from either long bimolecular RNA duplexes produced by RNAdependent RNA polymerases (RDRs) or extended hairpins and function in several epigenetic and posttranscriptional silencing systems (8, 9). Two siRNA-generating pathways have been described: (i) an endogenous siRNA-generating pathway that requires RDR2 and triggers changes in the chromatin state of elements from which they derive (10) and (ii) an exogenous siRNA-generating pathway that requires RDR6 and targets viral RNAs (3). Despite the difference in the origin and generation of their precursors, both siRNAs and miRNAs require DICER proteins for processing, and both are assembled into the RNA-induced silencing complex (RISC) to target their complementary RNAs (3).Extensive efforts are being made to identify regulatory noncoding small RNA sequences, primarily focusing on miRNAs. Many miRNAs have been detected by using an established cloning strategy (11-14), although the low abundance and tissue-specific expression patterns of some make experimental identification difficult. Thus, different computational approaches have been developed to predict miRNA sequences based on t...