Fig. 3: DFR charge controls target recognition efficiency on naked DNA and slow 1D diffusion.
a Left, scheme of a single-molecule experiment to detect Sox TFs binding to DNA. bt: biotin, nt: neutravidin; Right, microscopy images showing DNA loci in the AF647 emission channel (top, red), and bound Sox2 in the JF549 emission channel (bottom, yellow). Scale bar: 2 μm. b EMSA of indicated TFs with Sox motif-containing DNA. c EMSA at different NaCl concentrations for indicated TFs with Sox motif-containing DNA. d DNA templates of different lengths are used for single-molecule binding experiments. e Fluorescence time-trace of Sox2 (2 nM), binding to 50 bp motif-containing DNA, detected by JF549 emission (blue) and fitted with a step-detection algorithm (red). f Fluorescence time-traces of Sox TFs binding to DNA of indicated lengths. The raw data (gray) were fitted (blue, red). g Cumulative distributions of search times for Sox TFs, on various DNA substrates. Curve shading: SD (N = 4 for Sox2 with 50 bp DNA+; N = 3 for all the other cases). h Scheme of the 1D target search model: initially, a TF binds nonspecifically to DNA with ka. Once bound, the TF slides along the DNA with a speed of D1D for ατd until it dissociates from DNA. Upon reaching the target site, the TF either recognizes the target with kr, or bypasses it. sL: sliding length. KR: target recognition constant. i Specific on-rate curve as a function of DNA length. Fit: 1D target search model. Symbol: mean. Cap: SD (N = 4 for Sox2 with 50 bp DNA+; N = 3 for all the other cases). j-n Fitting and derived parameters from the fitting in i (see the “Methods” section): target recognition probability during 1D sliding (j) target recognition constant (k) target recognition rate (l) sliding length (m) and 1D diffusion coefficient (n). Line: mean. Cap: fitting error (N = 4 for Sox2 with 50 bp DNA+; N = 3 for all the other cases).
