Promoters recognized by forkhead proteins exist for individual 21U-RNAs

Abstract

C. elegans 21U-RNAs are equivalent to the piRNAs discovered in other metazoans and have important roles in gametogenesis and transposon control. The biogenesis and molecular function of 21U-RNAs and piRNAs are poorly understood. Here, we demonstrate that transcription of each 21U-RNA is regulated separately through a conserved upstream DNA motif. We use genomic analysis to show that this motif is associated with low nucleosome occupancy, a characteristic of many promoters that drive expression of protein-coding genes, and that RNA polymerase II is localized to this nucleosome-depleted region. We establish that the most conserved 8-mer sequence in the upstream region of 21U-RNAs, CTGTTTCA, is absolutely required for their individual expression. Furthermore, we demonstrate that the 8-mer is specifically recognized by Forkhead family (FKH) transcription factors and that 21U-RNA expression is diminished in several FKH mutants. Our results suggest that thousands of small noncoding transcription units are regulated by FKH proteins.

Figure 1. 21U-RNA-rich loci on chromosome IV are depleted of nucleosomes

(A) Abundance of 21U-RNAs (Batista et al., 2008) (bars) (left y-axis) at the indicated regions of chromosome IV and correlation plots (lines, right y-axis) between 21U-RNA loci and histone H3 enrichment peaks obtained by ChIP-chip (green), germline-specific histone H2B::GFP ChIP-chip enrichment peaks (dark yellow), nucleosome-protected regions in micrococcal nuclease digest of chromatin (Valouev et al., 2008) (blue) and predicted nucleosome occupancy based on underlying DNA sequence (black); see also Figure S1. (B) Correlation plots (lines, right y-axis) between 21U-RNA loci and histone H3 enrichment peaks obtained by ChIP-chip performed on young adult animals with germline (green), lacking germline - glp-4(bn2)- (orange) or lacking 21U-RNAs – prg-1(tm872)- (purple). All the correlations in 21U-RNA rich regions are statistically significant (p<0.05).

Figure 2. Nucleosome-depleted regions exist upstream of individual 21U-RNA loci

(A) Nucleosome occupancy per base pair averaged across all 21U-RNA loci and shown 100 nucleotides (top) or 500 nucleotides (bottom) upstream and downstream of the first U in 21U-RNA. The position of 21U-RNA (+1 to 21) on the X-axis is indicated in red, position of the conserved upstream motif (−59 to −25) is indicated in orange. (B) Average A+T content plots corresponding to regions shown in (A).

Figure 3. RNA Polymerase II is localized to the nucleosome-depleted regions upstream of 21U-RNA-producing loci

(A) Nucleosome occupancy per base pair averaged across all 21U-RNA loci and shown 500 nucleotides upstream and downstream of the first U in 21U-RNA is shown in blue and the Pol II occupancy per base pair averaged across all 21U-RNA loci is shown in red. (B) ChIP-qPCR with 8WG16 anti-Pol II antibody shows a significant Pol II enrichment over 21U-3016 and 21U-11909 loci in wild type adult worms, but not in glp-4 mutant worms lacking a germline compared to the IgG control, see Figure S2 for additional control genes. Error bars represent standard deviation from the mean from three independent experiments.

Figure 4. The CTGTTTCA 8-mer sequence upstream 21U-3372 is required for its expression

(A) Schematic of a 5kb region on chromosome IV indicating: positions of 21U-RNAs (vertical bars with arrows showing direction of their expression), deletion niDf199 present in the JU258 strain, and fosmid WRM0611aH08 used for complementation of the deletion. Expression of 21U-RNAs colored in red and green has been detected by RT-qPCR (B, C); the CTGTTTCA motif upstream of 21U-3372 (red) has been deleted in C. Additional annotated 21U-RNAs overlapping with those shown in (A) are omitted for clarity. (B, C) Quantification of 21U-RNA expression levels by RT-qPCR in the wild type N2 strain (blue), the JU258 strain with niDf199 (red), the transgenic strain carrying a fosmid covering the region deleted in JU258 (green), and the transgenic strain containing the fosmid with a deletion of the 8nt conserved motif next to 21U-R-3372 (orange). RNA was extracted from adult worms. Expression levels shown are relative to the N2 strain. miR-52 expression was used as an internal control. Error bars represent standard deviation from the mean from three independent experiments. See Figure S3 for additional control experiments. (D) qPCR on MNase-digested chromatin of the transgenic strain carrying a fosmid covering the region deleted in JU258 (green) and the transgenic strain containing the fosmid with a deletion of the 8nt conserved motif next to 21UR- 3372 (orange). The schematic indicates primers used to detect the upstream or downstream regions of 21U-3372 and 21U-11909. Error bars represent standard deviation from the mean from two independent experiments. See Figure S3 for MNase-digested chromatin preparations.

Figure 5. Transcription of the precursor for 21U-3372 starts 2bp upstream of the first T while the CTGTTTCA upstream sequence represents a DNA motif bound by nuclear proteins

(A) (left panel) Schematic representation of a 21U-3372 locus showing the position of the RT primer used for the detection of a precursor transcript and the transcription start site 2bp upstream of the 21U-RNA sequence. (right panel) Results of a 5′ RACE experiment showing the detection of a band corresponding to a 21U-RNA precursor on an agarose gel. * Indicates non-specific amplification. The sequence of 21U-3372 (red) and its precursor are shown. See Figure S4 for additional 5′ RACE experiments that suggest a lack of transcription through the upstream motif. (B) Quantification by RT-qPCR of the level of 21U-3372 precursor expression in young adult animals with germline (WT) and lacking germline, glp-4 (no germline). Error bars represent standard deviation from the mean from two independent experiments. (C) Electrophoretic mobility gel shift assay (EMSA) with nuclear protein extract from wild type adult worms and a 26bp dsDNA biotinylated oligonucleotide probe surrounding the 8-mer CTGTTTCA motif upstream of 21U-3372. The specificity of the binding was determined by incubating the reaction with 200x excess of non-biotinylated 26bp probe as a specific competitor or with 200x non-biotinylated 26bp probe with a mutated 8-mer motif (CGCCCGCA). Western blots on the right using anti-histone H3 and anti-actin antibodies show the nuclear (NUCL) and cytoplasmic (CYT) protein fractions prepared for EMSA. As expected, histone H3 is enriched in the nuclear fraction and actin in the cytoplasmic fraction. See Figure S5 for additional EMSA control experiments.

Figure 6. Germline-enriched Forkhead proteins bind to the CTGTTTCA 8-mer sequence and promote 21U-RNA production

(A, B) Quantification of the expression of 21U-RNAs by RT-qPCR in the unc-130(ev505) mutant (A, left panel) relative to WT, and upon simultaneous inactivation by RNAi of fkh-3, fkh-4 and fkh-5 (B, left panel) in the RNAi-sensitized eri-1 mutant background relative to control RNAi. RNA was extracted from L4 worms (A) or by picking 10 individual adult worms (B). The expression of 21U-RNAs was normalized to miR-52. Error bars represent standard deviation from the mean from at least three independent experiments and the number of asterisks indicates the p values as follows: * < 0.05, ** < 0.001, *** < 0.0001. mRNA quantification by RT-qPCR for prg-1 and unc-130 (A, right panel) or prg- 1 and fkh-3, fkh-4 and fkh-5 (B, right panel) is shown for the experiments presented in A and B (left panels). act-3 mRNA was used as an internal control. Error bars represent standard deviation from the mean from at least two biological replicates. In addition, PRG-1 protein levels in WT, unc-130 (ev505) and glp-4 (no germline) worms are shown by western (A, right panel). Anti-actin western was used as a loading control. See Figure S6 for 21U-RNA expression quantification in daf-16, lin-31 and fkh-9 mutants or pha-4(RNAi). (C) ChIP-qPCR detects specific enrichment of UNC-130 at the 21U-RNA loci. ChIP-qPCR was performed using anti-GFP antibodies in transgenic strains generated by bombardment and expressing UNC- 130::GFP from the germline promoter mex-5 (green bars). A transgenic line with no UNC- 130::GFP germline expression was used as control (grey bars). The enrichment of UNC- 130::GFP at different 21U-RNA loci was calculated relative to the control line. Three chromosomal regions were used as controls: a chromosome IV region located between individual 21U-RNA genes, the promoter region of the prg-1 gene and the coding sequence of ribosomal RNA (18S). Error bars represent standard deviation from the mean from three biological replicates and asterisk indicates p values < 0.05. Similar results were obtained with a MosSCI transgenic UNC-130::GFP line, see Figure S6. (D) Electrophoretic mobility gel shift assays (EMSA) with immunopurified bacterial recombinant FKH proteins fused with N-terminal Maltose Binding Protein (MBP) and a 26bp dsDNA biotinylated oligonucleotide probe surrounding the 8-mer CTGTTTCA motif upstream of 21U-3372. The specificity of the binding was determined by incubating the reaction with 200x excess non-biotinylated 26bp probe as a specific competitor or with 200x non-biotinylated 26bp probe with a mutated 8-mer motif (CGCCCGCA). See Figure S5 for additional EMSA control experiments.

Figure 7. Nucleosome-depleted promoters recognized by Forkhead proteins exist for individual 21U-RNAs

Schematic model of the 21U-RNA promoter.