Skip to main content

Structure and function of repeat-derived non-coding RNAs in epigenetic regulation of mammalian pericentric heterochromatin

Periodic Reporting for period 1 - RNA-SHAPE (Structure and function of repeat-derived non-coding RNAs in epigenetic regulation of mammalian pericentric heterochromatin)

Reporting period: 2015-05-01 to 2017-04-30

A significant fraction of mammalian genomes is composed of repetitive elements, such as pericentric major satellite repeats. Repetitive elements present a critical challenge to genome stability because these regions are prone to recombination. To safeguard genome integrity, these repetitive sequences are maintained in an epigenetically silent heterochromatic state to prevent mutagenic recombination. A confounding enigma revealed by recent transcriptome-wide studies is that repetitive elements are actively transcribed into noncoding RNAs (ncRNAs). Furthermore, these repeat-derived transcripts are themselves important to maintain the heterochromatic structure at repetitive elements. Nevertheless, it appears that a certain balance must be achieved since overexpression of repeat-derived ncRNAs is a hallmark of the most common types of cancer. How repeat-derived ncRNAs function in epigenetic regulation of heterochromatin is poorly understood. In this proposed study, ncRNAs transcribed from mouse pericentric major satellite repeats will be investigated as a paradigm to understand the precise mechanism of how these ncRNAs regulate heterochromatin formation. The key element of the proposed study is characterizing the function of major satellite RNAs from a structural perspective. This is a novel and promising approach because the functions of RNA molecules are intimately linked with their ability to fold into complex structures. A possible scenario is that accurate folding of major satellite RNA directs the local recruitment of chromatin complexes to pericentric loci. To address this hypothesis, the proposed research programme aims to achieve the following: (i) determine the structure of major satellite RNAs; (ii) determine if major satellite RNAs are divided into functional domains and how this relates to the structure; and (iii) determine how the structure of major satellite RNAs influence their interaction with pericentric DNA and chromatin-associated proteins.
In silico RNA structure prediction was performed using the RNAstructure Version 5.7 software. This showed that forward and reverse major satellite repeat RNAs (MSR RNAs) fold into distinct secondary structures. The forward transcript has more characteristic ring and hairpin structures whereas the reverse transcript has a more open single-stranded conformation with few stem loops.
To have an experimentally-derived RNA structural mode, in vitro RNA structure probing by SHAPE was initially done using SDS-PAGE gels to assay for radioactive primer extension products. However, quantification of SHAPE modifications using this approach proved to be difficult and the resolution is rather low. To address these issues, I shifted to assaying SHAPE modified RNA nucleotides by capillary electrophoresis (SHAPE-CE) using fluorescently labeled oligos in primer extension reactions. SHAPE modifications were quantified using the QuSHAPE software and was limit the constraints in SHAPE-directed RNA structure modeling. This was performed in collaboration with Fethullah Karabiber from Yilidiz University in Istanbul. In contrast to in silico predictions, in vitro SHAPE-CE showed the forward and reverse major satellite RNAs have strikingly similar secondary structures. The 5’ and 3’ ends of both transcripts come together to form a long-range stem. This extends into a central ring structure from which several radial stem loops emanate. The similar structures suggest that both forward and reverse MSR RNAs function in a similar fashion. The availability of single-stranded RNA stretches in the structure suggests that it can form RNA-DNA hybrids with major satellite DNA that are important for locus targeting. Moreover, the stem loops can mediate interactions of chromatin proteins with MSR RNA. Hence, MSR RNA could potentially function as a hub for RNA-protein interactions that are targeted to pericentric heterochromatin, which regulates the expression of major satellite repeats.
RNA structure probing in vivo
In vivo SHAPE structure probing was attempted in mouse embryonic stem cells (mESCs) using a variety of approaches to enrich for major satellite transcripts. However, the amount of MSR RNAs in cells is not abundant enough to be effectively assayed using radioactive methods. A possible solution is to use fluorescence-based methods for increased sensitivity, which is currently underway.
In collaboration with Howard Chang at Stanford University, we also looked at the transcriptome-wide RNA SHAPE-seq data from mESCs. However, as with the abovementioned limitation, the depth of sequencing reads over MSR RNAs is not enough to generate a reliable SHAPE-modification profile. Methods to improve SHAPE probing of low-abundance transcripts are currently underway in the Chang lab.
Investigation of RNA-DNA hybrids in vivo
As a complementary approach to investigate the architecture of major satellite transcripts, I checked for their ability to form RNA-DNA hybrids in vivo. Previous results from our lab indicated that treatment of extracted total nucleic acids with RNAse H, an enzyme that digests RNA in RNA-DNA hybrids, reduces the levels of MSR RNAs.
To learn more about the formation of RNA-DNA hybrids in vivo, I further studied endogenous RNAse H enzymes in mouse embryonic fibroblast (MEF) cell lines. There are two RNAse H enzymes in mouse, RNAse H1 and H2. I deleted Rnaseh1 in MEFs by CRISPR-Cas9 genome editing technologies and I also used Rnaseh2 null MEFs provided by Andrew Jackson from the university of Ediburgh. I also generated a double Rnaseh null cell line. Single deletion of Rnaseh1 or Rnaseh2 showed an accumulation of MSR RNAs, as assayed by RT-qPCR. This accumulation was even more significant in Rnaseh double null MEFs. These results indicated that endogenous RNAse H enzymes in mouse contribute to the processing of RNA-DNA hybrids formed by MSR-RNAs.
Another approached I employed to investigate RNA-DNA hybrids formed by MSR-RNAs is to target RNAse H to pericentric heterochromatin by TALEs recognize major satellite repeat sequences. This method is currently underway.
Investigating the protein interactome of MSR RNAs by ChIRP-MS
Chromatin isolation by RNA purification (ChIRP-MS) was attempted to elucidate chromatin proteins that bind to MSR RNAs. After optimization of sonication conditions to solubilize chromatin, forward and reverse MSR-RNAs were successfully enriched using this approach. However, analysis of bound proteins using SDS-PAGE did not show any detectable protein band. This is likely due to the issue raised above regarding the low abundance of major satellite transcripts in vivo. This method is currently being optimized.
The secondary structure models of forward and reverse MSR RNAs generated by SHAPE-directed RNAs structure probing strongly supports the potential for RNA-DNA hybrid formation. My investigation to follow up on this observation by studying mouse Rnaseh1 and Rnaseh2 has potential medical implications. Mutations encoding these enzymes result to an autoimmune disorder in humans called Aicardi-Goutières syndrome (AGS). The accumulation of RNA-DNA hybrids when Rnaseh is mutated leads to genome instability, thereby eliciting DNA damage and autoimmune responses in the brain. Dissecting the contribution of RNA-DNA hybrids formed by MSR-RNAs could further home in into the molecular mechanism of this autoimmune response, which could in turn be a target for therapeutics in the future.