Functional genetic screens interrogating biological pathways are a common tool in biomedical research. Targeted transcriptional silencing of specific genes often reveals their functions. Such screens allow researchers to remove some biases and explore vast regions of biology. Commonly screens identify exciting new therapeutic targets or previously uncharacterized cellular mechanisms that dominate an article. In fact, the design and output of a screen plays a massive role in which compounds or genetic elements are identified. Recent screening efforts in the Bedalov Lab (Clinical Research Division) demonstrated that the right design can reveal much about cellular mechanisms. This collaborative study with Fred Hutch scientists, Drs. William Grady and Patrick Paddison, was published in PNAS. In the article postdoc Smitha Sripathy and colleagues searched for pathways that silence a specific X-linked gene, MeCP2, to find that inactivation of genes on the X chromosome appears to be all-or-nothing – and that may not be a problem.
MeCP2 is an X-linked methyl-CpG-binding protein that is important in the development of neurons. Mutation of a single copy of MeCP2 results in severe cognitive impairment classified as Rett syndrome. Rett syndrome presents at young ages during neuronal development and occurs almost exclusively in females. Males with mutations in MeCP2 fail to reach full term or die in infancy because they have a single X chromosome. While females possess two copies of the X chromosomes, usually one of them is randomly silenced so no proteins are produced from its genes. In the case of Rett syndrome, individuals have a healthy copy of MeCP2, but it is turned off. Thus, the Bedalov Lab developed a screen to identify ways to reactivate the silent MeCP2 gene.
Image provided by Dr. Antonio Bedalov
Previous work with transgenic mice found that fusing a copy of MeCP2 to both firefly luciferase and the hygromycin resistance gene (MeCP2-FL-HR) resulted in loss of MeCP2 activity and Rett syndrome-like symptoms. While MeCP2 in this fusion was not functional, both the luciferase and hygromycin resistance proteins were active in tissues where MeCP2 is usually expressed. Fibroblasts were collected from female transgenic mice and two clonal populations were isolated – one where MeCP2-FL-HR was on the inactive X chromosome (Xi-8) and one where it was on the active X chromosome (Xa-3). These cells allowed the researchers to develop a screen for reactivation of the silenced MeCP2 gene. Under normal circumstances the Xi-8 cells died in hygromycin, but if an alteration activates the X chromosome near the MeCP2 gene, MeCP2, and consequently hygromycin resistance would allow them to survive. This approach was first validated using a DNA demethylating drug (5-AZA) known to broadly activate many regions of silent DNA, and with this established, proceeded to perform an shRNA screen for MeCP2 reactivation. The screen targeted 25,000 genes for RNA silencing, to determine whether reduced expression led to MeCP2 transcription allowing the cells to grow in hygromycin. Seven major classes of proteins were identified using this approach, including signaling kinases (Erb4, Pi3kcb), TGF-β proteins (Smad2, Bmpr2), Mitotic Kinases (Aurka, Plk2), and others. The screen also identified XIST, a molecule known to play a major and essential role in inactivating an entire X chromosome. Depletion of these targets were repeated for validation, during which researchers found the MeCP2 reporter signal was dramatically enhanced by treating with low levels of 5-AZA. This finding suggests multiple screen hits may need to be targeted for robust X chromosome activation. “We envision that efficient reactivation will require interrupting multiple pathways at the same time. This is not unlike approach that is pursuing in many disease states including cancer or common conditions such as hypertension where several drugs are use to achieve therapeutic benefit” said Dr. Antonio Bedalov.
The design of this screen was such that MeCP2 must be re-activated for cells to survive, but it did not exclude reactivation of the entire X chromosome. The scientists used the same library on a similar cell line containing a GFP gene on the inactive X chromosome and isolated cells that began expressing GFP. This counter screen identified at least four of the targets seen in the MeCP2 specific screen. The simplest explanation for this chromosome-wide inactivation is that the pathways identified activate XIST in some way. TGF-β proteins were some of the targets identified in both screens, and consistent with this hypothesis when each of six of these genes was depleted by shRNA, XIST RNA levels also decreased. One of these genes was BMP2 Receptor, which binds BMP2 and stimulates the rest of the pathway. Thus, when cells were exposed to BMP2 for 24 hours there was a dramatic increase in XIST RNA levels. The regulation of XIST RNA levels was also validated in vivo using transgenic mouse livers.
This study identified a previously unknown link between TGF-β signaling and X chromosome inactivation. However, the screen leaves many questions about therapeutic potential – a reasonable hypothesis being that doubling the amount of every gene on the X chromosome would have negative effects. The Bedalov Lab is currently testing this with promising preliminary results, “We are now testing reactivation in vivo in mouse brain. As XIST was such a prominent hit in the screen, we are testing whether deletion of XIST in neurons, after X inactivation has been completed, can reactivate MeCP2. Our initial results look promising: it seems we can obtain reactivation and animals tolerate it well” said Dr. Bedalov.
Funding for this research was provided by the Rett Syndrome Research Trust.
Sripathy S, Leko V, Adrianse RL, Loe T, Foss EJ, Dalrymple E, Lao U, Gatbonton-Schwager T, Carter KT, Payer B, Paddison PJ, Grady WM, Lee JT, Bartolomei MS, Bedalov A. 2017. Screen for reactivation of MeCP2 on the inactive X chromosome identifies the BMP/TGF-beta superfamily as a regulator of XIST expression. Proc Natl Acad Sci USA. 114(7):1619-1624. doi: 10.1073/pnas.1621356114