Investigation of bacterial nucleotide excision repair using single-molecule techniques

Abstract

Despite three decades of biochemical and structural analysis of the prokaryotic nucleotide excision repair (NER) system, many intriguing questions remain with regard to how the UvrA, UvrB, and UvrC proteins detect, verify and remove a wide range of DNA lesions. Single-molecule techniques have begun to allow more detailed understanding of the kinetics and action mechanism of this complex process. This article reviews how atomic force microscopy and fluorescence microscopy have captured new glimpses of how these proteins work together to mediate NER.

Keywords: Bacterial nucleotide excision repair; Single molecule; UvrA; UvrB; UvrC; UvrD.

Fig. 1

Prokaryotic nucleotide excision repair. Structural model of prokaryotic NER showing the key protein and steps in global genomic repair (GGR) and transcription coupled repair (TCR). TCR damage recognition is initiated by a stalled RNAP (PDB ID: 4LJZ) that recruits MFD (PDB ID: 2EYQ). MFD displaces RNAP and brings UvrA to the damaged site. In GGR, the UvrA₂B₂ complex (PDB ID: 3UWX) for the contact interface: 3FPN first searches for the distortion along the DNA caused by the lesion. Both pathways converge after the initial recognition steps. UvrA then transfers the damaged DNA to UvrB for damage verification. The dimeric UvrA protein (PDB ID: 2R6F) hydrolyzes both ATP and GTP. It also forms a complex with UvrB (PDB ID: 2FDC) and activates the ATPase activity of UvrB. During damage verification, the β-hairpin of UvrB (shown in turquoise) inserts between the two strands of DNA and forms a stable pre-incision complex, which is believed to activate UvrB’s ATPase. Binding and hydrolysis of ATP by UvrB is essential for recruitment of UvrC. The N-terminal endonuclease domain of UvrC (PDB ID: 1YCZ) initiates the cut 4–5 nucleotides 3′ to the damaged site followed by the 5′ cut by C-terminal endonuclease domain of UvrC (PDB ID: 2NRR) eight nucleotides away from the lesion. UvrD (PDB ID: 2IS1) unwinds the DNA and releases the oligonucleotide containing the lesion. Simultaneously, DNA polymerase I (PDB ID: 2HHQ) synthesizes the missing strand. Finally, DNA ligase I (PDB ID: 1DGS) seals the repair patch. All protein structures in this figure, with the exception of UvrB, are shown with a transparent surface and in ribbon presentation. UvrB is shown with its surface in orange for domains 1 to 3, and the β-hairpin is shown in cyan. C-ter, carboxy terminal; N-ter, amino terminal. From [6] with permission.

Fig. 2

Oxidized bases recognized by the UvrABC system. While 8-oxo-dG (OG) is a poor substrate for the UvrABC system, further oxidation products of this adduct are good substrates [55]. These include, guanidinohydantoin (Gh) and the two diastereomers of spiroiminodihydantoin (Sp), and spiroiminiohydantoin-adducts: Sp-lys, Sp-GPRP, Sp-GlcN, and Sp-GPRPGP. The F, is a fluorescein-modified thymine which serves as a positive control. Numbers in boxes indicate the extent of incision of a DNA duplex containing these site-specific lesions. Adapted from [55] with permission.

Fig. 3

A single ribonucleotide is a robust substrate for the UvrABC system. Electrostatic surface for 2′-0H of the ribose moiety embedded in a DNA duplex. The red spot around the O2′ indicates the negative electrostatic potential. Prepared by Yuqin Cai, NYU See reference: Yuqin Cai, Nicholas E. Geacintov, Suse Broyde Ribonucleotides as nucleotide excision repair substrates. DNA Repair 13 (2014) 55–60.

Fig. 4

UvrAB complex with UvrB-Qdot. A hemagglutinin (HA) epitope tag (YPYD-VPDYA) was engineered on to the N-terminus of UvrB to which was conjugated a mouse monoclonal antibody. A goat antimouse coated Qdot was bound to the UvrB-HA-Ab to make an “antibody” sandwich. This UvrB-HA-Ab-Ab-Qdot complex was mixed with UvrA and a 517 bp fragment prepared by PCR and containing a nick 40% from one end. This AFM image shows the transient complex of UvrA loading UvrB-HA-Ab-Ab-Qdot at the site of a nick, only 6% of the total complexes found on DNA had both UvrB and UvrA bound at the site of the nick. Adapted from [44] with permission.

Fig. 5

Nature of UvrAB movement on nondamaged DNA. UvrA shows no mobility once bound to DNA. Addition of UvrB to UvrA resulted in longer lived complexes with an average lifetime of 40 s. About 17% of the complexes showed motility on DNA exhibiting a range of motions including one-dimensional diffusion, directed motion that was ATP-dependent, and paused motion on the same DNA molecule following by rapid excursions to a new position on the DNA, hopping.

Fig. 6

Role of UvrB motifs in UvrBC movement. To conjugate UvrC to Qdots, the biotin ligase recognition sequence GLNDIFEAQKIEWHEGGG (AviTag™) was fused to the C-terminus of Bacillus caldotenax UvrC. At 50 mM KCl UvrC alone showed avid DNA binding, but no DNA sliding. Panel A. Addition of WT, or one of several mutant UvrB: Y96A, beta-hairpin deletion (Δhairpin) or D338N, resulted in DNA sliding. Panel B. 3D density plots of the diffusion constant versus the alpha factor for each UvrBC mutant complex. The coloring is a percentage scale relative to the maximum bin size. Panel C. UvrB–DNA co-crystal (PDB, 2FDC) Cartoon model with transparent surface of UvrB (green); non-damaged DNA strand (red); damage containing DNA strand (blue); beta-hairpin (magenta). Y96 and D338 show as space fill. Adapted from [46] with permission.