Extended Data Fig. 4: Genome-wide analysis of LFY-UFO binding.
a, Western Blot after DNA elution during ampDAP-seq experiment. After DNA elution, 20 µL of 1X SDS-PAGE Protein Sample Buffer was added to the remaining beads to run WB. Each lane represents one replicate. b, Assessment of experimental reproducibility of ampDAP-seq experiment through the comparison of replicates datasets 2 by 2. c, Effect of the LFY KARA mutation (K303A-R233A)51 on pAP3 activation in Arabidopsis protoplasts. Data represent averages of independent biological replicates and are presented as mean ± SD, each dot representing one biological replicate (n = 4). Unpaired t-tests (**: p < 0.01; ****: p < 0.0001). d, The LFY KARA mutation (K303A-R233A) does not disrupt LFY-UFO interaction in Yeast-Two-Hybrid (Y2H). EV = Empty Vector. LFY-40 is a LFY version lacking the first 40 aa and better tolerated by yeast cells. Values correspond to the different dilutions (OD = 7, 0.7 and 0.07). Top picture corresponds to the non-selective plate lacking Leucine and Tryptophan (SD -L-W), and bottom picture to the selective plate lacking Leucine, Tryptophan, Histidine and Adenine (SD -L-W-A-H). Pictures were taken at day + 4. e, Receiver operating characteristics (ROC) curves for mLUBS, dLUBS and LFY using the top 20% high-CFC LFY-UFO-specific peaks. Area under the curve (AUC) values are shown. TPR: True Positive Rate, FPR: False Positive Rate. f, Score distribution of LFY PWM with dependencies68 within dLUBS (best site on 20% most LFY-UFO-specific genomic regions, high CFC, n = 3843 genomic regions) and in canonical LFYBS (best site on 20% most LFY-specific genomic regions, low CFC, n = 3843 genomic regions). Best sites were selected within ±25 bp around the peak maximum. Wilcoxon rank sum test (****: p < 0.0001). Median (solid line), interquartile range (box edges), ±1.5 × interquartile range (whiskers) and outliers (black dot) are shown. g, De novo identification of URM from LFY ChIP-seq data25. Motifs identified at a fixed distance from LFY canonical binding sites in 298 regions harboring high LFY ChIP-seq to LFY ampDAP-seq coverage ratio. The text above each motif gives the motif’s start position relative to the canonical LFYBS, its length and the number of sites used to build the motif. h, EMSA with mLUBS and dLUBS highest score sequences. 6xHis-LFY-DBD is recombinant. UFO* refers to either recombinant ASK1-UFO-3xFLAG complex (top gel) or in vitro produced UFO-3xFLAG (bottom gel). Drawings represent the different types of complexes involving LFY-DBD (pale blue) and ASK1-UFO (red) on DNA. LFY-DBD binds as a monomer as previously reported29. The fact that in vitro produced UFO-3xFLAG shifts DNA in the presence of LFY indicates that ASK1 is not required for the UFO-LFY-DNA complex formation in vitro. i, EMSA with DNA probes corresponding to pAP1 and pAP3 DEE LFYBS and indicated proteins. Note that probes used here have the same length as those used to study LUBS. Source data are available in Supplementary Data 4.
