Science Spotlight

Getting all your DUX4 in a row

DUX4 expression observed in nuclei of iDUX4 myoblasts after doxycycline induction.
Figure provided by Dr. Sujatha Jagannathan

A major challenge for scientists is to successfully model human disease and biology in the laboratory. Each question may require a unique system to answer it – from proteins in a tube to transgenic mice, the model needs to recapitulate biology yet also allow for experimental manipulations. Recently, these questions have been engaging scientists studying facioscapulohumeral dystrophy (FSHD). FSHD is a form of muscular dystrophy caused by the aberrant expression of the transcription factor DUX4 in skeletal muscle leading to cell death and muscle dysfunction. One of the first cell culture models of this disease required differentiating FSHD myoblasts into muscle cells that express native DUX4. Recent advances in molecular biology have allowed researchers to control expression of DUX4 (and thus experimentally manipulate the gene); however, it has been unclear if these new cell culture models of FSHD successfully recapitulate the biology of this disease. In an article published in Human Molecular Genetics Dr. Sujatha Jagannathan and Sean Shadle in the Laboratories of Dr. Stephen Tapscott (Human Biology and Clinical Research Divisions) and Dr. Robert Bradley (Basic Science and Public Health Divisions) validate the new FSHD models by comparing RNA expression profiles.

The first developed cell culture model for FSHD used myoblasts. These less specialized cells divide in culture and with the right protein signals will differentiate into post-mitotic skeletal muscle. The differentiated cells aberrantly express the endogenous of DUX4 (enDUX), resulting in the activation of germline specific genes and errors in RNA quality control. Recently, researchers in the field instead delivered a constitutively expressed, non-native copy of DUX4 into myoblasts using lentivirus. Myoblasts expressing constitutive non-endogenous DUX4 (vDUX4) could only be studied for about 48 hours before dying. Moreover, a routine comparison between the RNA expression profiles of the enDUX4 and vDUX4 system showed significant differences.

Scientists in the Tapscott lab have expanded the repertoire of FSHD models by modifying myoblasts to contain a non-endogenous doxycycline-inducible DUX4 (iDUX4). These myoblasts contain this extra copy of DUX4 but are viable because they do not express it until doxycycline is present. Once doxycycline was added to iDUX4 cells they expressed DUX4, incorrectly expressed the key players previously identified in the enDUX4 cells, and shortly after died. The iDUX4 system gives researchers much more experimental control that is particularly useful for studying RNA expression profiles, because "The iDUX4 cells allow for synchronized temporal expression of DUX4, while the enDUX4 system is a mixed bag for when DUX4 expression is activated. Without that temporal control we get a very noisy dataset." said Dr. Jagannathan.

A cursory comparison of RNA expression profiles between all three systems identified the same key pathways and changes, but the devil is in the details. Previous studies comparing vDUX4 and enDUX4 found the large changes in RNA expression were similar but each system had many uniquely altered genes. Researchers measured the RNA expression profiles of iDUX4 cells and then performed a careful analysis between all three systems. Initial analysis revealed a similar trend; however, it appeared this might be due to 'edge effects', where RNAs expressed near the lower detection limit look dramatically different from sample to sample because of technical challenges rather than biology. When only RNAs expressed well above detection limits were analyzed the datasets became largely concordant. Eventually, it became clear the major differences between the three systems were due to basal expression of the specific cell type. Muscle specific genes were uniquely overrepresented in the enDUX4 system because those cells are completely differentiated. Innate immune response genes were activated in the vDUX4 system due to lentiviral delivery 48 hours prior to RNA profiling. Aside from these processes, all three systems gave consistent results.

Scientists in the Tapscott lab are very excited about what they can achieve with the new iDUX4 system. Sean Shadle explained, "Why DUX4 expression kills cells is an unsolved mystery in the field. It will help to answer that question with this new tool." They are taking multiple approaches to answer this question, the first of which is more RNA profiling. "Since the inducible cells go from perfectly healthy to completely dead in about 16 hours we can analyze [RNA] expression profiles at earlier time points and that might give us a clue about which pathways are involved." said Dr. Jagannathan. In a second approach, Shadle employs large scale genetic screens. "We are using this inducible DUX4 system to perform an siRNA screen for targets that prevent cell death to determine the pathways that are intrinsic to [DUX4-induced] cell death" said Dr. Tapscott. In the near future this expression system and the results it makes possible may lead to therapies for FSHD patients.

Funding for this research was provided by the National Institutes of Health, the Ellison Medical Foundation, Friends of FSH Research, and the FSH Society.

Jagannathan S, Shadle S, Resnick R, Snider L, Tawil RN, van der Maarel SM, Bradley RK, Tapscott SJ. 2016. Model systems of DUX4 expression recapitulate the transcriptional profile of FSHD cells. Hum Mol Genet. Epub ahead of print.