Chemical profiling of DNA G-quadruplex-interacting proteins in live cells

The intricate interplay between proteins and nucleic acids is crucial for regulating various cellular processes, including gene expression, DNA replication, and repair. Understanding these interactions within the native chromatin context requires robust methods. Chromatin immunoprecipitation (ChIP) coupled with mass spectrometry (MS)-based proteomics can characterize chromatin-associated protein complexes, but it requires high-affinity antibodies and typically analyzes one protein at a time. Enzyme-catalyzed proximity labeling techniques, like BioID and APEX, offer alternatives but suffer from slow kinetics, toxicity, and size limitations of fusion proteins. Photoactivation of small-molecule crosslinkers offers better spatial and temporal resolution, but existing affinity-based methods primarily map direct protein interactors of drugs or small molecules, not broader interaction networks. DNA G-quadruplexes (G4s) are non-canonical four-stranded nucleic acid structures found in human cells, dynamically forming and playing roles in gene regulation. While *in vitro* studies have identified some G4-interacting proteins, these studies lack the context of the native chromatin environment. This study aimed to develop a novel strategy to overcome these limitations and provide a more comprehensive understanding of protein interactions at specific functional genomic sites, focusing on G4s.

The literature highlights the importance of understanding protein-nucleic acid interactions in cellular processes. Existing methods, such as ChIP-seq coupled with mass spectrometry, have limitations in terms of scale and requirement for specific antibodies. Proximity labeling methods offer alternatives but suffer from issues like slow kinetics and potential toxicity. Existing studies on G4s have primarily focused on *in vitro* analyses or used methods that may not fully capture the dynamic nature of G4-protein interactions in a native chromatin context. This study builds upon previous work showcasing the potential of small molecules to target and stabilize G4s and the use of photocrosslinking for studying protein-protein interactions. However, it represents a significant advance by combining these approaches to profile G4-interacting proteins directly in live cells, offering a more comprehensive and holistic view of this complex interaction network.

This study developed a co-binding-mediated protein profiling (CMPP) approach to investigate DNA G4-interacting proteins in living cells. The method utilizes functionalized small-molecule ligands, based on the G4-selective ligand pyridostatin (PDS), that bind G4 structures in cellular chromatin. These ligands are modified with a photoreactive diazirine group for crosslinking and a click alkyne handle for subsequent labeling. Upon UV irradiation, the probes crosslink to nearby G4-interacting proteins. Two probes, photoPDS-1 and photoPDS-2, were synthesized, differing in linker length. A control probe lacking the G4-binding moiety was also included. The probes' G4-binding affinity and selectivity were assessed *in vitro* using fluorescence resonance energy transfer melting and fluorescence quench binding assays. *In vitro* proof-of-concept experiments used a G4-specific antibody (BG4) to demonstrate the efficiency of photoproximity crosslinking. For cellular studies, human HEK293T cells were treated with the probes, followed by photocrosslinking, nuclear protein extraction, and labeling with TAMRA-azide or biotin-azide for gel-based analysis or MS-based proteomics, respectively. Proteins significantly enriched in samples treated with photoPDS-1 or photoPDS-2 compared to the control probe were identified as candidate G4-interacting proteins. Selected candidates were further characterized *in vitro* using affinity enrichment coupled with western blot analysis and ELISA to assess G4 binding properties. Finally, the interaction of one candidate, SMARCA4, with endogenous G4s in chromatin was validated using ChIP-seq in K562 cells.

The CMPP approach successfully identified hundreds of putative G4-interacting proteins in human cells. *In vitro* assays confirmed high G4-binding affinity and selectivity for several of these novel candidates. The two probes used, photoPDS-1 and photoPDS-2, showed largely overlapping results, indicating that the linker length did not drastically affect the results. The identified proteins belonged to diverse functional classes, with a significant number involved in transcription, consistent with the established role of G4s in transcriptional regulation. A representative set of novel candidates—SMARCA4, UHRF1, RBM22, TTF2, DDX24, DDX1, and HMGB2—were selected for further validation. Six out of seven displayed G4-specific binding in affinity-enrichment experiments, with striking selectivity for different G4 topologies. ELISA further confirmed the high-affinity and selective G4 binding of SMARCA4, UHRF1, DDX1, DDX24, and RBM22. ChIP-seq analysis in K562 cells demonstrated that SMARCA4 binds to endogenous G4s in chromatin, particularly at gene promoters (42% of peaks), suggesting a role in promoter activity. This co-localization of SMARCA4 with endogenous G4s supports direct binding, and is not observed with sites that only have the potential to form G4s. The enrichment of the SMARCA4 signal is centered on the endogenous G4 sites. This supports the direct binding of SMARCA4 to the folded G4 secondary structure, and not to the underlying G-rich DNA primary sequence.

The findings demonstrate that the CMPP method successfully identifies genuine G4-interacting proteins in their native chromatin environment. The identification of hundreds of proteins, including known and novel G4-interactors, highlights the complexity of the G4 interactome and its involvement in various cellular processes. The validation of novel G4 interactors using orthogonal methods strengthens the reliability of the CMPP approach. The ChIP-seq data provide strong evidence of direct interaction between SMARCA4 and endogenous G4s in chromatin. The method’s ability to capture transient interactions, its broad applicability to various cell types, and its potential for mapping other nucleic acid structural features make it a valuable tool for future studies.

This study presents a novel chemical strategy, CMPP, for identifying cellular proteins that interact with G4 DNA structures in live cells. The method successfully identified hundreds of G4-interacting proteins, validated several novel candidates, and provided evidence for the direct interaction of SMARCA4 with genomic G4s in chromatin. The approach holds significant potential for future investigations into the roles of G4s in gene regulation and other cellular processes.

The study acknowledges that the G4 ligands used could potentially influence the endogenous G4 landscape and interactome. While short treatment times were employed to minimize this effect, the possibility of some perturbation cannot be entirely excluded. The study also notes potential differences in G4 interactomes between different cell lines, and the use of a different approach might lead to different results.