Our new TALE fusion is much more modular and easy to customize than the old zinc finger (Table 1). TALE construction is not heavily patented like zinc-finger design, and TALE’s are considerably cheaper to produce (Sanjana 2012 and Zhang 2011). They have made the zinc-fingers nearly obsolete as a tool for genetic engineering, including the creation of transgenic animal models (Tesson 2011, Sander 2011, Huang 2011, and Zhang 2011). We expect our new fusion protein could be translated to enable the creation of differentially methylated animal models, which would be revolutionary for epigenetic disease research (Klose 2006). In effect, this could transform epigenetics from a largely observational discipline to one of active intervention and manipulation, similar to the transition from early classical genetics to genetic engineering and synthetic biology.

We also fused a methyltransferase to a dCas9 DNA binding domain from the CRISPR-Cas system. We modified our MaGellin vector to include a cassette that expresses the necessary guiding sgRNA1. The sgRNA can easily be swapped out to quickly re-target the Cas system, much more quickly and with less synthesis than it would take to re-target the TALE. It has demonstrated activity but we did not have time to characterize it as completely as the TALE (Figure 6). We are intrigued by the prospect of optimizing the Cas system, which is considered to possibly have many advantages over the TALE (Gaj 2013). It has already been fused to transcription activators that similarly target promoters, so the strategy is theoretically sound (Bikard 2013). Particularly, it’s ability to multiplex and target multiple sites for methylation would be very useful (Wang 2013). However, many recent papers have described problems with Cas specificity that seriously limit its usefulness, so we expect the excitement around Cas may dim as scientists begin a rigorous re-evaluation of the comparison between Cas and TALE systems. We propose MaGellin as a simple, fast, and inexpensive assay to aid the development of superior Cas systems, as they are optimized by site-directed mutagenesis, directed evolution, or other methods. As outlined above and demonstrated with our TALE (Figure 5), fusing a methyltransferase to the Cas serves as a detectable marker for its DNA binding.

It is worth noting that the TALE-methyltransferase can be delivered to mammalian cells by adeno-associated vectors (AAV). This viral vector is the best currently available in terms of safety and efficiency (Daya 2008). Different serotypes have different cell tropisms, which can provide efficient cell-type targeting when used in conjunction with a cell-specific promoter (Ellis 2013). As we consider different model organisms, we may also want to try different methyltransferases. Importantly, the TALE, at 2.5kb long, can easily be packaged with a methyltransferase in an AAV vector, whereas a Cas would not fit (Konermann 2013).

We were not the only synthetic biologists developing epigenetic engineering tools this summer. In August, we were excited to see George Church’s lab at Harvard had fused TALEs to histone modifiers, enabling another distinct form of epigenetic engineering (Konermann 2013). They delivered these TALE fusions to mammalian cells with AAV, which gives us confidence our system can be translated into more complex organisms. With their targeted histone modification and our targeted DNA methylation, it appears the era of serious epigenetic engineering efforts is now upon us.