Supplementary Figure 3: CRE-DOG appears to act independently of protein splicing.

(a-b) Luciferase assays testing various CRE-DOG^OG variants for GFP-dependent activation of a CALNL-luc2 reporter. (a) Mutations known to disrupt the protein splicing ability of VMA intein did not drastically alter CRE-DOG^OG activity. N-splicing site refer to a cysteine residue in the N-intein portions of N-CreintG of CRE-DOG^OG. C-splicing site refer to a cysteine residue in the C-intein portions of C-CreintG of CRE-DOG^OG. n=6 or 9 replicate samples each condition. (b) Both intein portions were required for the tight GFP-dependent recombination activity of CRE-DOG^OG. “Int” indicates presence of intein portion. Removal of either intein portion adversely affected CRE-DOG^OG specificity for GFP. Each portion independently contributed to the GFP-specificity of CRE-DOG^OG. n=9 each condition. In box plots, boxes show 25^th to 75^th percentile, line in box shows median, whiskers show minimum to maximum range. (c) Testing for protein splicing mechanism for CRE-DOG^OG function. FLAG tagged C-terminal component of CRE-DOG^OG is ~52 kDa. If protein splicing occurs, then the ~37 kDa full-length Cre product should be detected by anti-FLAG. (d) Western blot testing the protein splicing model in (c). 293T cells transfected with indicated components (CRE-DOG^OG-FLAG and GFP) were lysed 1 day post-transfection. Lysate supernatant (nuclear and cytoplasmic content) were probed for protein products labeled by FLAG tag. The results failed to support a protein splicing mechanism, due to absence of ~37 kDa band. Consistent results were obtained in an independent experiment with whole cell lysate loading.

Source data