Identification of phylogenetically conserved sequence motifs in microRNA 5' flanking sites from C. elegans and C. briggsae

Abstract

Background: MicroRNAs (miRNAs) are small, noncoding RNA molecules that act as post-transcriptional regulators of gene expression. Studies concerning transcriptional regulation of miRNAs have so far concentrated on those located within the intergenic region of the genome and the search for putative promoters, thus leaving open the question of the existence of possible regulatory elements common to all miRNAs including those located in introns of protein coding genes.

Results: In this study, we initially searched for motifs occurring in the area 1000 bp upstream from all miRNAs independent of their genomic location. We discovered a previously unknown sequence motif GANNNNGA that displayed a conserved distribution in the nematode worms Caenorhabditis elegans and Caenorhabditis briggsae. This motif had a peak occurrence at 500 bp upstream, with a sharp drop-off toward the miRNA start site. Further analysis indicated that this motif was locally restricted and not enriched 1000-5000 bp upstream or 0-2000 bp downstream of the miRNA start site. In addition, this motif was observed to be most abundant in the upstream sequences of two important miRNAs, mir-1 and mir-124. This abundance was also conserved in phylogenetically distant species including human and mouse.

Conclusion: The results show that the motif GANNNNGA is conserved close to miRNA precursor start sites, suggesting that it may be involved in miRNA sequence recognition or regulation. This data provides important knowledge for the identification and computational prediction of miRNA sequences.

Figure 1

The main workflow of this study.

Figure 2

Significant conserved motifs discovered from the miRNA regulatory region of C. elegans and C. briggsae. Presented is the total number of the motif instances in both regulatory region sequence sets and the number of miRNAs that contain the motif in their upstream region. The rightmost column shows the p-value of the motif calculated against random background sequences. Motif logos show the information content present at every position in the sequence, for example in the logo of motif GANNNNGA the G:s and A:s are certain, thus having the highest information content (2 bits) while in the positions 3 to 6 all four nucleotides have about the same frequencies and the information content in these positions is close to zero.

Figure 3

The frequency distribution of motif GANNNNGA around the miRNAs of C. elegans (a) and C. briggsae (b). All occurrences of the motif GANNNNGA were located from the sequences 0–5000 bp upstream and 0–2000 bp downstream of all C. elegans and C. briggsae miRNA start sites and the corresponding 200 bp central moving average curves were drawn for both worms.

Figure 4

The frequency diagrams of the motif GANNNNGA occurrences in sequences 1000 bp upstream of mir-1 (panel a) and mir-124 (panel b) family members in C. elegans, C. briggsae, human and mouse. The mir-1-1 and mir-124-a1 1000 bp upstream sequences of human and mouse contain similar number of occurrences of the motif GANNNNGA as the corresponding sequences in C. elegans and C. briggsae.

Figure 5

miRNA microarray expression analysis in C. elegans. Dots represent the absolute expression level (vertical axis) of wildtype N2 C. elegans miRNAs from whole L4 stage animals as a function of the number of GANNNNGA motifs (horizontal axis). The line indicates a regression plot of the data (r² = 0.13).