... notes that Fgf8 is kept silent by near-locus wide Polycomb binding, which spreads into your insertion region. It might be that Fgf8's Polycomb spreads across the MusD element, preventing DNA methylation and H3K9me3 deposition. So it is silent as long as Fgf8 is, and...
... when Fgf8 is activated, the whole locus loses Polycomb, and MusD finds itself derepressed. If this is true, that means that it is not the Fgf8 enhancers that control its promoter but just its Polycomb repression, which is in sync with that of Fgf8. This is all speculation, of course.
Thank you for your comments and thoughts! In the paper we tested whether the MusD will respond to enhancers at 3 other developmental loci and it does! So we don’t think it is related to Fgf8 landscape but rather the MusD promoter itself
Also, in Bl6 mice, the MusD is methylated and silenced despite being inserted at the same position (figure 1 and reported previously in Kano et al.)- this is caused by a polymorphism between mouse strains (what exactly is still under investigation 😉)
Quite possible - but the repression could still be Polycomb-mediated, especially since you notice that the MusD demethylation is constitutive there. And many developmental loci have Polycomb marks that spread widely from the target gene itself.