Mapping transcription factor-DNA interactions using femtosecond laser cross-linking and nanopore sequencing.

• Applicant: Professor Dr. Christoph Russmann
• Subject Area: Biophysics ; General Genetics and Functional Genome Biology ; Structural Biology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 538668510

Every cell in an organism has the same set of genetic information. Cells need only part of this information at any one time to maintain basic functions. Most of it is only required under certain conditions or in certain cell types. The regulation of genes occurs mainly at the transcriptional level with the help of transcription factors (TFs). TF-DNA binding is mediated by weak hydrogen bonds that easily escape biochemical analysis. Crosslinking can convert these bonds into strong covalent bonds. Chemical crosslinking, although efficient, disturbs the balance in the cell. Crosslinking with UV lamps is inefficient and damages the DNA to such an extent that analysis is difficult. Therefore, femtosecond laser crosslinking (FLIX) was developed to efficiently crosslink TF to DNA with little DNA damage. Nanopore sequencing (NPS) is a third-generation sequencing method. Inside the pore, the DNA molecule occupies a volume that partially restricts ion flow, which is observed as a drop in ion current. Depending on the geometry, size, and chemical composition of the molecule, the intensity of the ion current and the duration of the translocation change. The resulting raw signal - called a "squiggle" - is translated into the nucleotide sequence by "basecalling". NPS enables sequencing with high accuracy, read length, and throughput, as well as the detection of base modifications such as DNA methylation. In the proposed project, FLIX and NPS will be combined to study TF-DNA interactions at the level of single nucleobases with amino acid residues. Thus, structural biological questions about TF-DNA interactions will be answered in a highly specific, sensitive, and fast way by direct "readout" using NPS. The central goal is the detection of specific nucleobase-amino acid complexes of a TF-DNA interaction. For this purpose, the specific, previously unknown "squiggles" of the complexes have to be translated into nucleotide sequences by the "basecalling" process. Starting with unirradiated and irradiated double-stranded DNA fragments, double-stranded DNA fragments after cross-linking with different TFs, the "basecalling" algorithms of a convolutional neural network (CNN) are trained to assign the novel "squiggles" to nucleobase-amino acid complexes and DNA damage. By varying the irradiation conditions, different mono- and biphotonic DNA damages can be generated. The different DNA modifications are detected by biochemical and physical methods. The feasibility of the method will be demonstrated in vitro and in vivo using the DNA binding activity of the transcription factors TBP and NF1, cell culture experiments, and ChIP-Seq. The gold standard will be FLIX-based mass spectroscopy.

• DFG Programme: Research Grants