Supplementary Figure 9: Choice of polymerase to maximize C-to-T signals during PCR.

(a) 76-mer oligonucleotide with two 5fC sites labeled with AI, PCR amplified and then subjected to Sanger sequencing. The graph shows that the polymerase can readily read through the modified sequence even with two successive 5fC-AI adducts next to each other (in a 5ʹ-fCGfCG-3ʹ context). (b) 76-mer control oligonucleotide with G and A replacing the two 5fCs in the sequence in a. (c) Sanger sequencing results of Taq polymerase reading through the adduct between 5fC and AI. PCR bias will diminish the effective C-to-T signals; hence different polymerases were tested to maximize the signal for 5fC detection. Alternatively, new labeling probes could be explored to minimize chemical scarring on the 5fC base³¹. (d) Commercially available MightyAmp DNA polymerase best overcomes the PCR bias and hence allows maximal C-to-T transition during PCR. (e) Scheme diagram of the on-bead wash step (with NaOH solution) to remove any strands that do not contain 5fC (“supernatant” depicted in Step 2). Therefore, the output (depicted in Step 3) is enriched for 5fC-AI-SH. (f) AI-labeled 76-bp dsDNA with a central 5fC-AI-SH:G pair. The C-T transition peak can be observed at the 5fC site, although the signal of C (resulting from the G in the complementary strand) is higher than that of T (from 5fC-AI-SH). (g) Elution from the DTT-cleavage step. The eluted fraction is enriched for 5fC-AI-SH, and hence a near-complete C-to-T transition peak can be observed at the 5fC site.