Every day, human cells may experience up to 105 DNA damage events, making maintenance of genetic integrity critical for survival. Of the many types of DNA damage, double-stranded breaks (DSBs) are the most cytotoxic DNA lesions. RAD51-mediated DNA homologous recombination (HR) provides an error-free approach to repair DNA DSBs and is an important process for maintaining genetic integrity3. This recombination process also requires the coordinated actions of a plethora of protein cofactors and is transient and highly dynamic. Due to lack of structural information on key intermediates, protein cofactor complexes and conformational changes, the mechanistic details of this process remain poorly understood. This project will lead to a more detailed understanding of the mechanism of RAD51, the major protein involved in HR.

The major idea is to employ a novel RAD51-DNA mini-filament, developed in my laboratory, in structural investigations by single-particle and time-resolved cryoEM. The preliminary results reveal local structural variants of dsDNA bound to RAD51. Based on this promising outcome, we will use this approach to reveal the structure of additional key intermediates (such as the D-loop and RAD51 functional mutants) and RAD51–protein cofactor complexes. The long-term goal is to Couple this mini-filament with time-resolved sample preparation and allow monitoring of protein motion over the course of the RAD51-mediated HR process. Our mini-filament design is novel and has resulted in the first ternary complex of the protein cofactor bound to a RAD51–DNA filament which represents a breakthrough in the DNA recombinase field.

Different Rad51 nucleoprotein filament

Antimicrobial-resistant (AMR) bacteria caused 1.27 million deaths in 2019. Antibiotic overuse has accelerated the emergence of AMR bacteria. Phages are viruses that infect bacteria and outnumber bacteria by about a factor of ten. Due to urgent threats to AMR bacteria, there has been renewed interest in phage therapy. Advanced engineered phage, to enhance specificity and infectivity, have been used to successfully cure patients with infected lung diseases. Although the overall architecture of tailed mycobacteriophages looks similar, the dramatic genome divergence prevents sequence homolog search and phage engineering by sequence approach. Due to the limited knowledge of the molecular architecture of phages at the atomic resolution, the current engineering processes only rely on random mutation across the phage genome followed by the selection of effective phages, which will take years.

Structurally, tailed phages can be divided into four regions: capsid, neck (head-to-tail interface), tail, and baseplate (tip). By cryo-EM and divided-and-conquer approach, we recently solved the entire tailed phage from siphoviridae (the most abundant tailed phages) structure where head, neck, tail, and baseplate reach to 3.2, 3.3, 2.2, and 3.0 Å resolution, respectively (the highest resolution among reported entire phage structure of siphoviridae, to our knowledge). At this resolution, we can immediately observe something novel: 1) a small peptide protein stabilizes the capsid packed with viral genome; 2) the portal ring and the tape measurement protein function as the viral genome gating; 3) the origin of the tail tube flexibility; 4) the tape measurement protein forms a homotrimer and interact with the baseplate. Although overall structure of siphoviridae looks similar, the sub-region can be very different. For example, our preliminary cryo-EM screening already shows three different molecular architectures of the baseplate from our mycobacteriophage collection. 

Combining with the phage library we collected in Taiwan, we are now in the position to reveal a repository of phage structures, enabling comprehensive sequence-and-structure analyses to study functional convergence without sequence similarity and to pinpoint conserved structural modules for phage infectability. We also apply advanced cryo-ET technology to study phage-host interactions.

Protein arginine methyltransferases (PRMTs) play roles in cancer progression by methylating many cancer related proteins. However, the question remains which protein methylation leads to oncogenic activation. Our structural approaches focusing on PRMT-protein complex provide a way to answer this question and can assist inhibitor design to prevent methylation for cancer therapy.