Our Team’s Solution In order to address the challenges associated with developing new site-specific methyltransferase proteins, our team proposed several different strategies.
Proposed Solution. First, we proposed a migration away from mammalian systems and into E. coli. E. coli does not have a native cytosine methyltransferase, and therefore offers a noise-free environment for methylation studies. Any methylation of CpG sites in E Coli would be a product of a candidate engineered protein rather than the native organism. Second, we envisioned a modular one-plasmid system that can be employed for quickly and cheaply screening the activity and specificity of any DNA binding domain – methyltransferase fusion protein. This plasmid-based methylation assay is called MaGellin.
Plasmid Features. To accommodate the MaGellin assay, our team designed a plasmid with several key features:

1. CpG Methyltransferase (M.SssI) with a generic linker sequence in the cloning site. For a working fusion protein and assay, only a DNA binding domain must be cloned into the plasmid. This inherently standardizes MaGellin and lessens the time a user of the assay must spend cloning.
2. Multiple cloning site downstream of T7 promoter for orthogonal expression of fusion protein in T7 Express competent E. coli.
3. Cloning site for a smaller DNA sequence, specific to the fusion protein being screened – named the “target site”, where the protein will bind. This can be the binding site for a CRISPR-Cas, TALE, Zinc Finger, or transcription factor
4. AvaI restriction site 4 bases downstream of the target site – the AvaI restriction enzyme is blocked by methylated CpG sites, thus screening for site specific methylation becomes equivalent to screening for AvaI digestion
5. AvaI restriction site sufficiently further downstream of the target site – named the off-target site. This site screens for non-specific DNA methylation as it is spatially removed from where the fusion protein binds to the plasmid.
6. XbaI site for linearization of the plasmid. Linearizing the plasmid simplifies analysis of the AvaI digestion by gel electrophoresis.
7. Validated bisulfite conversion primer binding sites, so users do not need to go through the time-consuming primer design process if they choose to fortify MaGellin’s results with bisulfite sequencing results
8. sgRNA cloning site for users who want to target a CRISPR-Cas binding domain. The sgRNA is constitutively expressed and can be swapped by restriction digest.
9. Validated bisulfite conversion primers for users who choose to advance to bisulfite sequencing, for even higher resolution in detecting methylation, after proving their enzyme’s efficacy with our MaGellin assay.
10. Kanamycin resistance as a selection marker

Noiseless Chassis. After cloning this plasmid, we faced the challenge of choosing the correct cell line for the assay. We chose to transform into T7 Express cells for several reasons: T7 RNA Polymerase in the lac operon allows us to turn on expression of fusion protein after induction with IPTG In the T7 Express cell line, genes for several restriction enzymes known to target methylated DNA are knocked out (McrA-, McrBC-, EcoBr-m-, Mrr-). This ensures that our assay plasmid is not cleaved in vivo. Results are difficult, if not impossible, to interpret in the commonly used BL21 cell line. MaGellin Workflow. The workflow for screening new fusion proteins with the one plasmid MaGellin bacterial system is as follows: Clone DNA binding domain into our backbone Transform into T7 Express cells Induce with 1mM IPTG for 5 hours + miniprep Digest using AvaI and XbaI Run on a gel and look for 3 distinct band patterns that correspond to specific and interpretable methylation outcomes. The presence of large one band corresponds to non-site-specific DNA methylation (AvaI was blocked at both the target and off target sites, and thus only XbaI cut the plasmid) The presence of two bands corresponds to site-specific DNA methylation (AvaI was only blocked at the target site, thus AvaI cut in the off target site and XbaI cut the plasmid) The presence of three bands corresponds to no DNA methylation – or an inactive fusion protein (AvaI was not blocked at either the target or off target sites and XbaI cut the plasmid) (Figure 2) Figure 2: A mock gel of the MaGellin assay for targeted methylation. Different methylation patterns are easily detected by distinct band patterns on the gel. Complementary Software. We realized early on that the MaGellin assay could easily lend itself to quantification, and we designed a software package to do just that. The MaGellin software accelerates experimental analysis and removes human bias. It is unique because its bioinformatics module can predict expected band lengths based on the methylation sensitivity of the restriction enzymes. Validating MaGellin With Methyltransferase. First, we tested MaGellin with a purified methyltransferase in vitro. The results made it clear that MaGellin can detect methylation, at both the “target” and “off-target” site. MaGellin is also sensitive to various degrees of methylation. These experiments helped us optimize the ideal amount of plasmid and restriction enzyme to use in any study moving forward. With a Validated Fusion. Now that we knew MaGellin could report DNA methylation in a quantifiable manner, we wanted to express an older fusion. We recreated the old zinc finger – methyltransferase fusion protein that had been previously described and assayed its activity and specificity. What we found using MaGellin was in agreement with the numerous studies that had focused on assaying this fusion protein. We found that overexpression of the zinc finger fusion protein led to many off target methylation events, confirming what others had noted about this fusion protein – that it is not suitable for site specific DNA methylation (Figure 3). THIS FIGURE NEEDS to be REMADE! Remember our discussion? Very confusing to follow that “Non Induced” actually means induced. Either show data from NEB5 cells, write ‘low induction v high induction’ or just compute the Off-to-On target ratio and report that. we need to hcnage this figure but it will be good. We need to have way more detail on the gel image obviously. We should be comparing zf with in vitro methylation Figure 3: Expression of a previously published zinc finger-methyltransferase fusion in E.coli (blah well write this caption when we add the actual gel)