An enhancer RNA recruits KMT2A to regulate transcription of Myb

Abstract

The Myb proto-oncogene encodes the transcription factor c-MYB, which is critical for hematopoiesis. Distant enhancers of Myb form a hub of interactions with the Myb promoter. We identified a long non-coding RNA (Myrlin) originating from the -81-kb murine Myb enhancer. Myrlin and Myb are coordinately regulated during erythroid differentiation. Myrlin TSS deletion using CRISPR-Cas9 reduced Myrlin and Myb expression and LDB1 complex occupancy at the Myb enhancers, compromising enhancer contacts and reducing RNA Pol II occupancy in the locus. In contrast, CRISPRi silencing of Myrlin left LDB1 and the Myb enhancer hub unperturbed, although Myrlin and Myb expressions were downregulated, decoupling transcription and chromatin looping. Myrlin interacts with the KMT2A/MLL1 complex. Myrlin CRISPRi compromised KMT2A occupancy in the Myb locus, decreasing CDK9 and RNA Pol II binding and resulting in Pol II pausing in the Myb first exon/intron. Thus, Myrlin directly participates in activating Myb transcription by recruiting KMT2A.

Keywords: CP: Molecular biology; H3K4me3; KMT2A; MLL1; Myb; Pol II pause release; enhancer RNA; enhancer hub; erythroid differentiation; gene expression; lncRNA.

Figure 1.. The Myb −81-kb enhancer is the TSS for the non-coding Myrlin RNA

(A) ENCODE ChIP-seq of LDB1 complex components (GATA1, LDB1, and TAL1), CTCF, and Pol II in the Myb-Hbs1l locus in MEL cells. PolyA RNA-seq is shown in MEL cells. Intergenic LDB1 complex and CTCF binding sites are highlighted (orange and blue vertical bars). (B) Nested PCR products from 5′ and 3′ RACE in MEL cells. M represents size marker. (C) Poly-A RNA-seq and Ldb1 ChIP-seq tracks from ENCODE for MEL at the −81-kb enhancer and downstream Myrlin exon 2. (D) Prediction of coding potential for Myrlin and select other transcripts as determined by CPAT (see text). (E) Relative expression of Myrlin in nuclear and cytoplasmic fractions of MEL cells determined by RT-PCR. MALAT1 lncRNA and ActB provided nuclear and cytoplasmic controls, respectively. (F) Total RNA of uninduced and induced MEL cells was used to determine relative expression of Myrlin and Myb by RT-PCR. Expression was normalized to ActB. (G) Total RNA of E8.5 yolk sac cells (YS) and E14.5 fetal liver cells (FL) was used to determine relative expression of Myrlin, Myb, and Hbb-b1 by RT-PCR. Expression was normalized to ActB. Error bars indicate SEM of 3 independent biological experiments. *p < 0.05; **p < 0.01; ***p < 0.001 by Student’s t test. See also Figures S1.

Figure 2.. CRISPR deletions at the TSS for Myrlin affect Myb expression

(A) Schematic diagram of the CRISPR-Cas9-mediated 17-bp deletion upstream of the Myrlin TSS. Smaller deletions were within the 17-bp deletion or extended upstream. PCR (below) shows the relative sizes for the 7-, 8-, and 17-bp deletion PCR products using wild-type or mutant gDNA. (B) Relative expression of Myrlin (exon 1, exon1/2 junction, exon 2) and Myb (exon 2) in MEL CRISPR-Cas9 control cells (no gRNA) and MEL 7-, 8-, and 17-bp CRISPR-Cas9 deletion mutants. Expression was normalized to ActB. (C) RNA Pol II-ChIP in MEL CRISPR-Cas9 control cells and MEL 17-bp CRISPR-Cas9 deletion mutant (ΔTSS) at the Myb gene promoter and enhancer sites (−36, −61, −68, −71, and −109 kb). (D) ChIP for TBP at the Myb enhancer sites as in (C) for MEL CRISPR-Cas9 control cells and MEL ΔTSS CRISPR-Cas9 deletion mutant. (E) Relative expression of Hbb-bh1, Hbb-b1, and Hbb-y in induced MEL CRISPR-Cas9 control cells and MEL ΔTSS CRISPR-Cas9 deletion mutant. Error bars indicate SEM of 3 independent biological experiments. *p < 0.05; **p < 0.01; ***p < 0.001 by Student’s t test. See also Figure S2.

Figure 3.. Chromatin organization of the Myb locus and transcription factor occupancy is affected by reduction of Myrlin in ΔTSS cells

(A) 3C interaction frequency between the enhancers found within the Myb-Hbs1l intergenic region using the Myb promoter as the anchor (black bar) observed for ΔTSS and control cell uninduced MEL cells. (B) GATA1 occupancy at Myb enhancers in control and ΔTSS uninduced MEL cell lines determined by ChIP-qPCR. (C) TAL1 occupancy at Myb enhancers in control and ΔTSS uninduced MEL cell lines determined by ChIP-qPCR. (D) LDB1 occupancy at Myb enhancers in control and ΔTSS uninduced MEL cell lines determined by ChIP-qPCR. (E) H3K27ac normalized to H3 occupancy at Myb enhancers in control and ΔTSS uninduced MEL cell lines determined by ChIP-qPCR. Error bars indicate SEM of 3 independent biological experiments. *p < 0.05 and **p < 0.01 by Student’s t test.

Figure 4.. CRISPRi targeting of Myrlin compromises Myb transcription but not hub formation

(A) Expression of Myb and Myrlin monitored by RT-qPCR in CRISPRi uninduced MEL cells targeted with dCas9 sgRNA1 or sgRNA3 or without an sgRNA (control). (B) ChIP-qPCR for H3K9me3 across the Myb locus before and after Myrlin CRISPRi in uninduced MEL cells. (C) ChIP-qPCR for H3K27ac across the Myb locus before and after Myrlin CRISPRi. (D) ChIP-qPCR for LDB1 across the Myb locus before and after Myrlin CRISPRi. (E) 3C interaction frequency between the Myb enhancers using the Myb promoter as the anchor (black bar) observed after Myrlin CRISPRi in uninduced MEL cells targeted with dCas9 sgRNA or without an sgRNA (control). Error bars indicate SEM of 3 independent biological experiments. *p < 0.05 and **p < 0.01 by Student’s t test.

Figure 5.. Myrlin interacts with KMT2A complex

(A) ChIP-qPCR for H3K4me3 across Myb sequences. (B) ChIP-qPCR for KMT2A at the Myb gene. (C) ChIP-qPCR for KMT2A complex component WDR5 at the Myb gene. (D) Myrlin RNA ChIP to detect interaction of Myrlin with KMT2A complex components WDR5, KTM2A, and menin. Glyceraldehyde 3-phosphate dehydrogenase served as control. (E) RNA pull-down using biotinylated Myrlin and blotting using antibodies to KMT2A components. TUB served as a negative control. Error bars indicate SEM of 3 independent biological experiments. *p < 0.05; **p < 0.01; ***p < 0.001 by Student’s t test. See also Figure S3.;

Figure 6.. Pol II and CDK9 occupancy within the Myb enhancer hub is affected by CRISPRi targeting of Myrlin

(A) ChIP-qPCR for CDK9 in Myrlin CRISPRi uninduced MEL cells targeted with dCas9-KRAB or without an sgRNA (control). (B) RNA Pol II occupancy in the Myb promoter/exon 1 region after Myrlin CRISPRi and in control cells. (C) CUT&Tag for RNA Pol II Ser5 and Ser2 phosphorylated forms. (D and E) Quantitation of CUT&Tag data showing Pol II Ser5P and Ser2P occupancy in the Myb gene body and at each of the enhancers. (F) Separate analysis of the shorter (<120 bp) and longer (>270 bp) Pol II Ser5P CUT&Tag fragments displaying Pol II Ser5 occupancy. Orange color represents regions of RNA Pol II occupancy shared in both control and CRISPRi conditions. Dotted line represents TSS. (G) Pausing index calculated for Pol II Ser5 across Myb. Error bars indicate SEM of 2 independent biological experiments. *p < 0.05 and **p < 0.01 by Student’s t test. See also Figure S4.

Figure 7.. KLF1 interaction with Myrlin contributes to localization within the Myb enhancer hub

(A) Myrlin RNA pull-down was conducted to determine the efficiency of probes for ChIRP. (B) ChIRP DNA pull-down by Myrlin across Myb and the Myb enhancers. (C) RNA ChIP for KLF1. (D) Biotinylated Myrlin pull-down and blotting with KLF1 antibodies. TUB served as a negative control. (F) Model of the Myb locus enhancer hub. Transcription of Myb is depicted with and without the −81-kb enhancer Myrlin eRNA after Myrlin CRISPRi. Large, shaded circle represents Pol II and LDB1 transcription factor density within the Myb enhancer hub, which is diminished when Myrlin transcription is reduced. Error bars indicate SEM of 3 independent biological experiments. *p < 0.05 and **p < 0.01 by Student’s t test.