DUX4 expressing cells R-N-A pickle

Science Spotlight

DUX4 expressing cells R-N-A pickle

From the Tapscott Lab, Human Biology Division

April 17, 2017

Biological studies often require scientists to dedicate large efforts to generating tools. Resources such as antibodies, chimeric proteins, and mouse or cell culture models of disease enable exciting studies but can be the biggest challenge for researchers. In a previous publication researchers in the Tapscott Lab (Human Biology Division) developed a new cell culture model for facioscapulohumeral dystrophy (FHSD), which is now being leveraged for further study. FHSD is a specific form of muscular dystrophy caused by the incorrect expression of a transcription factor, DUX4, in skeletal muscle. When active DUX4 is incorrectly expressed in cell culture or mice it leads to cell death; however, scientists in the field are still uncovering why DUX4 expression induces cell death. In their previous work the Tapscott Lab developed a system for doxycycline dependent expression of DUX4 (termed DUX4i). Prior to this, most studies delivered an active DUX4 gene into cells by viral means. Viral delivery stimulates an innate immune response in many cell types and the expression levels are often highly variable – both of which can be largely avoided using the DUX4i system. Sean Shadle, a graduate student in the Tapscott Lab, has used this system to perform an siRNA screen looking for pathways that cause cell death after DUX4 activation. The findings were recently published in PLOS Genetics.

FHSD is largely studied in precursor muscle cells called myoblasts. Fully differentiated muscle cells do not replicate and thus cannot be used for many experiments. Primary myoblasts also present challenges because they can only replicate in culture for a few weeks. For experiments that require longer periods of growth researchers can use rhabdomyosarcoma cells, cancer cells with similarities to myoblasts. Shadle and colleagues used the DUX4i system to generate a clonal DUX4 inducible rhabdomyosarcoma cell line. Similar to DUX4 expression in primary myoblasts, 48 hours after DUX4 induction 95% of these cells had died. The Fred Hutch Genomics Core helped to establish a high throughput screen to identify genes necessary for DUX4 toxicity. For the screen, DUX4-expressing cells were transfected with siRNAs targeting 6,961 genes within the ‘druggable’ human genome – each gene being targeted by a pool of 4 siRNA. From this large group of genes 69 siRNA pools were found to significantly improve survival after DUX4 expression. A subset of these were reproduced both using pools and the individual siRNA. The technically validated hits were then screened for their ability to inhibit doxycycline-dependent gene expression, this allowed researchers to eliminate targets that increased survival by turning off DUX4 expression rather than silencing a pathway downstream of DUX4.

DUX4 induces nuclear dsRNA accumulation and phosphorylation of EIF2AK2/PKR. Fluorescence microscopy using the J2 antibody to detect dsRNA in RD-DUX4i cells treated with doxycyline for 16 hours or transfected with poly(I:C) – a synthetic dsRNA. Note that not all nuclei showed dsRNA aggregation, likely due to asynchronous activation of the DUX4 transgene. Western blot of phosphorylated and unphosphorylated PKR in extracts from RD-DUX4i cells treated with doxycyline for 21 hours or transfected with poly(I:C).

Image provided by Sean Shadle

In the end, one gene that stood out was MYC. MYC is an important regulator in many diseases, it is thought to be up-regulated in FHSD, and its binding partner MAX also scored in the screen. As MYC is a key initiator of apoptotic pathways, this hit was not surprising. Though how MYC was stabilized proved intriguing. MYC protein and mRNA levels increased with DUX4 expression not because of increased RNA transcription, but rather because the mRNA molecules were stabilized. In fact, DUX4 expression seems to repress mRNA degradation in general, not just for MYC, as the lifetime of many mRNAs increased; however, the effect was particularly strong for MYC.

Finding that mRNA stability was generally increased in DUX4 expressing cells implicated another screen hit as particularly important. The gene RNASEL also scored in the screen and is a component of innate immune response to viral infection, which acts on viral RNA. Thus, it is possible that DUX4 increases the stability of cellular RNA to a point that stimulates an immune response and triggers apoptosis. Viral infection often stimulates an immune response by inducing high levels of double-stranded RNA (dsRNA), which are otherwise uncommon in cells. Consistent with this, DUX4 expressing cells had high levels of dsRNA and the active phosphorylated version of a dsRNA antagonist, PKR. While the content of the dsRNA foci were not tested in this study, the Tapscott Lab has plenty of hypotheses and plans to understand why they arise. Currently many possibilities exist – for example, long non-coding RNAs are often produced from the anti-sense strand of protein coding regions and thus could form duplexes. Shadle outlined other possibilities, “We are particularly excited about determining the nature of the dsRNA sequences induced via DUX4. For example, do they correspond to intramolecular dsRNA from inherently structured sequences activated by DUX4, or to intermolecular dsRNA from repetitive sequences induced by DUX4?”

Another interesting aspect of this disease is the particular effect on skeletal muscle. Performing the screen in the correct cell type is essential, but the question remains how these targets function in other tissue types. The Tapscott Lab is starting to answer some of these questions already. Said Shadle, “DUX4 is cytotoxic when expressed in a multitude of cell lines including 293T (kidney), HepG2 (liver), and human myoblasts and it is even toxic when expressed in mouse cells. Therefore, we would predict that the screen hits are not specific to muscle cells, though this is something that still remains to be tested. Currently, our understanding is that FSHD muscle is particularly susceptible to the misexpression of DUX4 rather than a muscle specific toxicity”. As with most studies, the answers found here only prompt more, exciting questions.

Funding for this research was provided by Friends of FSH Research, the National Institutes of Health, and FHCRC Genomics. 


Shadle SC, Zhong JW, Campbell AE, Conerly ML, Jagannathan S, Wong CJ, Morello TD, van der Maarel SM, Tapscott SJ. 2017. DUX4-induced dsRNA and MYC mRNA stabilization activate apoptotic pathways in human cell models of facioscapulohumeral dystrophy. PLoS Genet, 13(3), e1006658.