Jeannie Lee specializes in studying the role of long noncoding RNA (lncRNA) in epigenetic regulation and uses X chromosome inactivation as a model system. She is particularly interested in how lncRNAs interact with chromatin complexes to change gene expression. Her laboratory formulates paradigms in RNA biology and develops methodologies to probe interactions at the RNA-protein interface, with the long-term goal of translating knowledge of basic mechanisms to new therapeutic strategies.
X Chromosome Inactivation as a Model
X chromosome inactivation (XCI) equalizes gene expression between male (XY) and female (XX) mammals by silencing one X chromosome in the female embryo. In this way, genes are expressed from the two female X chromosomes at the same level as from the single male X chromosome. XCI is an excellent model in which to study long noncoding RNA (lncRNA) because the epigenetic process is controlled by the X inactivation center (Xic), a region on the X chromosome that harbors many lncRNAs. These transcripts interact with protein factors to control the initiation, spread, establishment, and maintenance of silencing on a 150-megabase scale. Some examples include the following:
- Xist RNA, a 17 kb transcript that coats the X chromosome and spreads the Polycomb repressive complex 2 (PRC2) throughout the X chromosome.
- Tsix RNA, a 40 kb antisense transcript that controls X chromosome pairing and Xist expression.
- Jpx RNA, a shorter transcript that activates Xist by evicting a the repressor transcription factor, CTCF.
- Xite, an enhancer-associated RNA that controls Tsix expression and X chromosome pairing.
- RepA RNA, a 1.4 kb repeat-rich RNA that targets PRC2 to the X inactivation center.
Many aspects of XCI remain poorly understood. Current lab projects in this arena focus on (1) the protein interactomes for Xist and Tsix RNA, (2) how Xist RNA spreads and targets silencing factors on a chromosome-wide scale, (3) how Tsix prevents this inactivation cascade, (4) how imprinting might be controlled by mechanisms involving transgenerational inheritance, and (5) how "counting" of chromosomes is effected. We suspect that RNA-protein interactions will be central to these problems.
Why Long Noncoding RNA?
Two properties of mammalian lncRNAs render them excellent vehicles by which to deliver epigenetic control:
1. Cis-regulation and allelic-specific control: long transcripts are naturally tethered to the site of synthesis via the act of being transcribed. Long noncoding RNAs can therefore function as allele-specific tags and offer the possibility of recruiting chromatin complexes in cis. This property may explain their prominence within the Xic and imprinted regions. Proteins do not retain allelic memory, as their transcriptional origin is lost when mRNA is shuttled to the cytoplasm for translation to protein.
2. Locus-specific targeting: long transcripts can also direct chromatin complexes to a unique location in the genome. Transcription factors (proteins) generally recognize short DNA motifs that occur thousands of times in the genome. They therefore regulate expression of multiple genes at once. RNA's potential for locus-specific targeting may account for why so much of the mammalian genome is transcribed.
Chromatin Modifiers, Their RNA Interactomes, and Disease
Long noncoding RNAs have emerged as key players in development of disease. For instance, we recently showed that loss of Xist RNA and X inactivation in somatic cells causes a highly lethal blood cancer. The pathway toward disease is of major interest. Many chromatin-associated factors are now known to interact with RNA. We would like to identify these networks of RNA-protein interactions using RNA-centric and protein-centric approaches. In an RNA-centric approach, we developed iDRiP-mass spec and, using Xist RNA as bait, we identified >100 proteins that interact with Xist RNA. This protein interactome includes many repressive epigenetic factors and also chromosome architectural factors. In protein-centric approaches, we have developed improved methods of performing CLIP-seq and RIP-seq for proteins such as PRC2 and CTCF. Both PRC2 and CTCF associate with thousands of transcripts. Members of these RNA-protein interactomes are candidate biomarkers and therapeutic targets in human disease. We would like to understand how such networks of RNA-protein interactions control normal development and progression toward disease. Such efforts will open up new ways of treating human diseases such as Rett, CDKL5, and Fragile X Syndromes.
This work is funded in part by the National Institutes of Health and the Rett Syndrome Research Trust.
As of April 25, 2016