biostudies-arrayexpressUnknownTranscriptomicsGenomicsProteomicsHarmen BussemakerotherDrosophila melanogasterDrosophila melanogasterhttps://www.ebi.ac.uk/biostudies/studies/E-GEOD-65073Binding of transcription factors to DNA is mediated by the recognition of the chemical signatures of the DNA bases and the three-dimensional shape of the DNA molecule. The direct contribution of DNA shape to DNA-binding specificity has been difficult to assess, as DNA shape is a consequence of its sequence. Here, we teased apart these two modes of recognition in the context of Hox-DNA binding. We made a series of mutations in Hox residues that, in a co-crystal structure, only recognize DNA shape, and tested the effect on DNA binding preferences using SELEX-seq. Analysis of shape features of selected sequences revealed that these residues are both necessary and sufficient for selection of sequences with distinct shape features. We used statistical machine learning to show that the accuracy of binding specificity predictions improves by adding shape features to a model that only depends on sequence. We conclude that shape readout is a direct and critical component of binding site selection by Hox proteins. Three rounds of SELEX were performed on a series of Hox mutants as described in Slattery et al, Cell, 2011 (PMID 22153072) and Riley et al, Methods in molecular Biology, 2014 (PMID 25151169). Briefly, His-tagged Scr and Antp mutant proteins were incubated with a randomized 16mer oligonucleotide library, and bound DNA was amplified and sequenced as described (PMID 22153072, PMID 25151169).biostudies-arrayexpressGrowth Protocol - All plasmids were transformed into BL21(DE3) (Agilent). 5ml overnight bacterial culture was diluted into 200mL (LB) and grown to OD 0.6. Protein expression was induced with 1mM IPTG for 4 hours at 37C.Library Construction - All His-tagged proteins were purified by cobalt chromatography (Clontech). Bradford assays and Coomassie staining was used to assess protein quality and concentration. Libraries were prepared according to methods described in Slattery et al, Cell, 2011 (PMID 22153072) and Riley et al, Methods in Molecular Biology, 2014 (PMID 25151169)OrganizationMINSEQE ScoreAssays and DataProcessed DataMAGE-TAB FilesRichard MannMatthew SlatteryHarmen BussemakerRemo RohsData Transformation - Illumina 2500 fastq files were demultiplexed according to barcode (according to Riley et al, Methods in Molecular Biology, 2014) For Hox-Exd complexes, 12/14/16-mer affinity tables were constructed from sample-specific Round 3 data together with Round 0 reference data using the method described in Riley et al, Methods in Molecular Biology, 2014 (PMID 25151169). For Hox monomers, 9-mers affinity tables were constructed from sample-specific Round 3 data (Round 2 for Dfd) together with Round 0 control data using the method described in Riley et al, Methods in Molecular Biology, 2014 (PMID 25151169). In the 16-mer affinity tables, the variable region (upper case) was extended with the fixed 5' and 3' flanking sequences of the library. Genome_build: Not applicable - SELEX reads are not aligned to reference genome Supplementary_files_format_and_content: List of 9/12/14/16-mer relative binding affinity tables derived from R2/R3 and R0 in tab-delimited formatfalseDeconvolving the recognition of DNA shape from sequenceBinding of transcription factors to DNA is mediated by the recognition of the chemical signatures of the DNA bases and the three-dimensional shape of the DNA molecule. The direct contribution of DNA shape to DNA-binding specificity has been difficult to assess, as DNA shape is a consequence of its sequence. Here, we teased apart these two modes of recognition in the context of Hox-DNA binding. We made a series of mutations in Hox residues that, in a co-crystal structure, only recognize DNA shape, and tested the effect on DNA binding preferences using SELEX-seq. Analysis of shape features of selected sequences revealed that these residues are both necessary and sufficient for selection of sequences with distinct shape features. We used statistical machine learning to show that the accuracy of binding specificity predictions improves by adding shape features to a model that only depends on sequence. We conclude that shape readout is a direct and critical component of binding site selection by Hox proteins. Three rounds of SELEX were performed on a series of Hox mutants as described in Slattery et al, Cell, 2011 (PMID 22153072) and Riley et al, Methods in molecular Biology, 2014 (PMID 25151169). Briefly, His-tagged Scr and Antp mutant proteins were incubated with a randomized 16mer oligonucleotide library, and bound DNA was amplified and sequenced as described (PMID 22153072, PMID 25151169).2015-01-21T00:00:00Z2023-09-11T22:33:56.124Z2021-10-04T11:33:00ZE-GEOD-65073SRP052569