1. Genome-wide Mapping of DROSHA Cleavage Sites on Primary MicroRNAs and Noncanonical Substrates 
Baekgyu Kim, Kyowon Jeong, V. Narry Kim 
Molecular Cell 
Volume 66, Issue 2, Pages e5 (April 2017)
DOI: /j.molcel
Copyright © 2017 Elsevier Inc. Terms and Conditions

2. Molecular Cell 2017 66, 258-269.e5DOI: (10.1016/j.molcel.2017.03.013)
Copyright © 2017 Elsevier Inc. Terms and Conditions

3. Figure 1 fCLIP-Seq Captures the Products of DROSHA with Intact Ends
(A) Workflow of DROSHA fCLIP-seq. See STAR Methods for details. Abbreviations: CIAP, calf intestinal alkaline phosphatase; PNK, polynucleotide kinase; IP, immunoprecipitation.
(B) Comparison of two technical replicate libraries from HEK293T cells. Each dot in the scatterplot represents log10 (RPKM, reads per kilobase per million) value of an miRNA gene (blue, n = 486) and a non-miRNA gene (gray, n = 34,733). R stands for Pearson’s coefficient between two libraries.
(C) Distribution of DROSHA fCLIP-seq reads around all annotated miRNA hairpin loci (n = 1,881) in HEK293T cells.
(D and E) DROSHA fCLIP-seq reads represent products from DROSHA-mediated processing. (D) Distribution of DROSHA fCLIP-seq reads and RNA-seq reads around the MIR106B cluster locus in HEK293T. (E) Top: gray bars indicate DROSHA fCLIP-seq reads aligned to the MIR106B locus. For visualization, 5% of fCLIP-seq reads were randomly sampled. Red and blue vertical lines indicate the start or end positions of the most mapped reads. The identified 5′ and 3′ cleavage sites are marked as red and blue arrowheads, respectively. Bottom: miRBase-annotated hairpin structure of pri-mir-106b. Mature 5p and 3p miRNAs are indicated in purple, and the positions of the miRNA ends and identified sites are indicated by arrowheads.
See also Figure S1.
Molecular Cell  , e5DOI: ( /j.molcel )
Copyright © 2017 Elsevier Inc. Terms and Conditions

4. Figure 2
Identification of DROSHA Cleavage Sites on Pri-miRNAs in HEK293T Cells
(A and B) The 5′ cleavage sites (left) and 3′ cleavage sites (right) identified in HEK293T cells (A) and HeLa cells (B). Note that “multiply identified sites” are individually counted. When an miRBase site is available, each site is classified either as “consistent with miRBase site” or “inconsistent with miRBase site.” If a miRBase site is missing on the given strand, the identified site is counted as a “newly defined site.”
(C) The discrepancy between “identified site” and “miRBase site” indicates mono-uridylation of pre-let-7i. The positions of the 3′ identified site and 3′ miRBase site are marked as red and purple arrows, respectively. Top left: hairpin structure of pri-let-7i. miRBase-annotated sequences of let-7i-5p and -3p are indicated in purple. Right: reads from DROSHA fCLIP-seq (gray bars) and reads from AGO-loaded small RNA-seq (skyblue bars) aligned to the 3′ strand of pri-let-7i. For presentation, 7.5% of fCLIP-seq reads and 1% of small-RNA-seq reads were randomly sampled. miRBase-annotated sequence of let-7i-3p is indicated in purple. The red and purple dashed lines show the 3′ identified site and 3′ miRBase site, respectively. Bottom left: schematic diagram of LET-7I biogenesis involving mono-uridylation.
See also Figure S2 and Tables S2 and S3.
Molecular Cell  , e5DOI: ( /j.molcel )
Copyright © 2017 Elsevier Inc. Terms and Conditions

5. Figure 3 Alternative Processing of Pri-miRNAs
(A) Histogram showing the number of pri-miRNAs with uniquely or multiply identified sites in HEK293T (left) and HeLa cells (right).
(B) Hairpin structures of mmu-Mir-342 (top) and hsa-MIR-342 (bottom) with their multiple cleavage sites. miRBase-annotated sequences of mature miRNAs are indicated in purple. Red and blue arrowheads indicate the positions of multiple cleavage sites from Wu et al. (2009) (top) and those identified from fCLIP-seq in HEK293T and HeLa cells (bottom). In each pri-miRNA, pairs of red and blue arrowheads indicate long and short pre-miRNAs, respectively. Sequence differences between two pri-miRNAs are marked in red on hsa-MIR-342.
(C) Hairpin structures of the other 11 pri-miRNAs with evidence for alternative cleavage on both arms of the hairpin in HEK293T cells. miRBase-annotated sequences of mature miRNAs are indicated in purple. The positions of multiply identified sites on each miRNA are marked as pairs of red and blue arrowheads that indicate long and short pre-miRNAs generated from the miRNA, respectively.
See also Figure S3 and Table S2.
Molecular Cell  , e5DOI: ( /j.molcel )
Copyright © 2017 Elsevier Inc. Terms and Conditions

6. Figure 4 Identification of Noncanonical DROSHA Substrates
(A) Genome-wide search in HEK293T (top) and HeLa (bottom) cells yields hairpins that show a clear cleavage pattern and the 2-nt 3′ overhang signature. “Junction” indicates the hairpins spanning two different genomic locations such as exon-intron or UTR-coding sequence (CDS).
(B) Cleavage pattern detected on the ANKS6 mRNA. Top: hairpin structure folded from the sequences around the identified cleavage sites (black arrowheads). Bottom: reads from DROSHA fCLIP-seq (gray bars) aligned to the ANKS6 hairpin. For presentation, 5% of reads from each library were randomly sampled. Vertical black lines indicate the identified cleavage sites.
(C) Proportion of the 5′ (left) and 3′ (right) ends of AGO-loaded small RNA reads at positions relative to the identified sites on noncanonical substrates. Only the identified sites with at least ten mapped small RNA reads within a ±10-nt window were considered.
(D) Abundance of AGO-loaded small RNA species (x axis). The y axis shows the number of genes mapped on noncanonical substrate loci (red), mirtron loci (green), or miRNA loci (blue).
See also Table S4.
Molecular Cell  , e5DOI: ( /j.molcel )
Copyright © 2017 Elsevier Inc. Terms and Conditions

7. Figure 5 DROSHA Cleavage-Mediated Gene Regulation
(A) Location of non-miRNA exonic hairpins (indicated in blue) chosen for validation. See Figure S4 for secondary structures and sequences of the substrates.
(B) In vitro processing of internally labeled pri-miR-16-1 hairpin, DGCR8 hairpin, and noncanonical substrate hairpins by recombinant DROSHA-DGCR8 complex. Cleavage products are indicated as F1, F2, and F3. The juxtaposed lanes separated by a dashed line are not contiguous, but all of them are from a single gel.
(C) qRT-PCR measurements of the exons containing DGCR8 hairpin and noncanonical substrates. Relative levels after DROSHA and DICER knockdown are shown, compared with control siRNA (small interfering RNA)-treated cells. The expression levels were normalized by GAPDH mRNA level. Error bars represent SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001, by two-sided two-sample t test, equal variance; n = 3.
(D) Cumulative plot showing the changes of mRNA levels measured by RNA-seq upon knockdown (KD) of DROSHA (top) and DICER (bottom). Exonic hairpin-hosting mRNAs (noncanonical substrates) are indicated with a red line, and all other mRNAs (background) are indicated with a black line. Out of 25 identified genes, 19 genes with at least ten mapped RNA-seq reads are included in this calculation. The same criteria were applied for the background genes.
See also Figures S4 and S5 and Table S4.
Molecular Cell  , e5DOI: ( /j.molcel )
Copyright © 2017 Elsevier Inc. Terms and Conditions