Journal club: Jaenisch lab paper on epigenetic CRISPR-Cas9 rescue of Fragile X in a dish

Liu, et al. Cell Fragile X
Liu, et al. Cell graphical abstract

There’s much more to CRISPR-Cas9 than just gene editing and a new paper from the lab of Rudy Jaenisch in Cell highlights that in an exciting way. It reports epigenetic reversal of a Fragile X Syndrome phenotype in induced pluripotent stem cell (IPSC) neurons.

Fragile X Syndrome is a neurological disorder in boys resulting from CGG repeat expansions in the regulatory region of the FMR1 gene and associated epigenetic alterations including DNA methylation that tend to shut off gene expression.

The new paper, Liu, et al., is entitled “Rescue of Fragile X Syndrome Neurons by DNA Methylation Editing of the FMR1 Gene.” The pub is not only technologically exciting, but also provides mechanistic insights into Fragile X, with some new insights into important DNA methylation aberrations.

The paper’s highlights section has four take-home points:

  • “Targeted demethylation of CGG repeats by dCas9-Tet1 reactivates FMR1 in FXS cells
  • Demethylation of CGG repeats induces an active chromatin status for FMR1 promoter
  • Methylation-edited FXS neurons behave similarly as wild-type neurons
  • FMR1 reactivation by dCas9-Tet1 is sustainable in a human/mouse chimeric model”

You can also see the graphical abstract above.

Liu, et al. Figure 6C Fragile X CRISPR, Cell
Liu, et al. Figure 6C Fragile X CRISPR, Cell

The dCas9-effector domain fusion with Tet1 reversed aberrant heterochromatin at the FMR1 gene and as a result the gene’s expression went up to more normal levels. This in turn led to physiological changes in neurons derived from patient IPSCs manifesting as more normal neuronal electrical activity. Strikingly, these phenotypic rescue was maintained in the human cells even in vivo in a chimeric mouse model.

dCas9 is a catalytically inactive form that when fused to Tet1 can lead to reductions in DNA methylation so no genetic change involved with this form of CRISPR-Cas9, but still a pronounced functional outcome.

From an MIT piece by Nicole Giese Rura on the paper:

“These results are quite surprising—this work produced almost a full restoration of wild type expression levels of the FMR1 gene,” says Jaenisch, who is a professor of biology at Massachusetts Institute of Technology and the senior author on the study. “Often when scientists test therapeutic interventions, they only achieve partial restoration, so these results are substantial.”

The reactivated FMR1 gene rescues neurons derived from fragile X syndrome induced pluripotent stem (iPS) cells, reversing the abnormal electrical activity associated with the syndrome. When rescued neurons were engrafted into the brains of mice, the FMR1 gene remained active in the neurons for at least three months, suggesting that the corrected methylation may be sustainable in the animal.”

There were some interesting details and technical aspects to this paper.

I was impressed that they did a lot of work to look for potential off-target effects of dCas9-Tet1 on DNA methylation and gene expression along with binding of dCas9-Tet1  elsewhere besides the FMR1 gene. There were some functional off-target effects, but reassuringly of much lower magnitude of activity than at the correct target locus. Such off-target effects can be reduced by using less dCas9-Tet1.

The team also compared results with the epigenetic approach to deletion of CGG repeats (see Figure 6C above). This yielded functionally similar results both at the level of neuron activity and DNA methylation, more tightly linking these mechanisms to Fragile X.

Overall, this is an exciting study and illustrates the power of combining different technologies with IPSC approaches. It also highlights again just how powerful IPSCs are more modeling human diseases in ways that would be otherwise difficult or impossible. Human IPSCs are particularly useful for neurological diseases where living cell samples are generally not available or expandable in culture for study. Many different diseases have already been modeled via IPSCs and even in one case researchers aimed to model pain in a dish.


  1. Mechanistically it is quite a surprising result, suggesting that gene expression is a direct result of DNA methylation, rather than DNA methylation being just a consequence of lack of expression. I am wondering for how many loci a similar approach would be successful. And can you do the opposite? Forcefully methylate a locus to shut down expression?

  2. Dear Paul,

    I just read a little bit more about the amazing article above. Of course just one case report, but an amazing one. The concept sounds really interesting. What do you think about it?

    see injectible bone grafts and pre designed bone grafts

Leave a Reply