LigD: A Structural Guide to the Multi-Tool of Bacterial Non-Homologous End Joining

Abstract

DNA double-strand breaks are the most lethal form of damage for living organisms. The non-homologous end joining (NHEJ) pathway can repair these breaks without the use of a DNA template, making it a critical repair mechanism when DNA is not replicating, but also a threat to genome integrity. NHEJ requires proteins to anchor the DNA double-strand break, recruit additional repair proteins, and then depending on the damage at the DNA ends, fill in nucleotide gaps or add or remove phosphate groups before final ligation. In eukaryotes, NHEJ uses a multitude of proteins to carry out processing and ligation of the DNA double-strand break. Bacterial NHEJ, though, accomplishes repair primarily with only two proteins-Ku and LigD. While Ku binds the initial break and recruits LigD, it is LigD that is the primary DNA end processing machinery. Up to three enzymatic domains reside within LigD, dependent on the bacterial species. These domains are a polymerase domain, to fill in nucleotide gaps with a preference for ribonucleotide addition; a phosphoesterase domain, to generate a 3'-hydroxyl DNA end; and the ligase domain, to seal the phosphodiester backbone. To date, there are no experimental structures of wild-type LigD, but there are x-ray and nuclear magnetic resonance structures of the individual enzymatic domains from different bacteria and archaea, along with structural predictions of wild-type LigD via AlphaFold. In this review, we will examine the structures of the independent domains of LigD from different bacterial species and the contributions these structures have made to understanding the NHEJ repair mechanism. We will then examine how the experimental structures of the individual LigD enzymatic domains combine with structural predictions of LigD from different bacterial species and postulate how LigD coordinates multiple enzymatic activities to carry out DNA double-strand break repair in bacteria.

Keywords: DNA double-strand break; LigD; ligase; non-homologous end joining; phosphoesterase; polymerase; protein structure and function.

FIGURE 1

Non-homologous end joining repair of a DNA double-strand break in bacteria. Ku homodimers bind the DSB, recruiting LigD to the DNA ends. The DNA ends can be processed by LigD to add nucleotides through the LigD polymerase domain. LigD can also remove ribonucleotides or convert a 3′-phosphate to a 3′-hydroxyl via the LigD phosphoesterase domain. Once the DNA ends consist of a 5′-phosphate and a 3′-hydroxyl, the LigD ligase domain seals the phosphodiester backbone, repairing the DSB. Image created with Biorender.com.

FIGURE 2

Domain arrangement and in silico models of Ku. (A) Domain arrangement of Ku in bacteria capable of non-homologous end joining. MIN, minimal C-terminus; EXT, extended C-terminus. (B) Comparison of conservation of sequence and length between the minimal and extended Ku C-terminus. (C–E) In silico models of (C) M. tuberculosis, (D) P. aeruginosa and (E) B. subtilis Ku homodimers, predicted by ColabFold (Mirdita et al., 2021). Purple, core domain; blue, minimal C-terminus; pink, extended C-terminus.

FIGURE 3

Domain arrangement of LigD in bacteria capable of non-homologous end joining. POL, polymerase domain; PE, phosphoesterase domain; LIGASE, ligase domain.

FIGURE 4

Atomic structures of the LigD ligase domain in open and closed conformations from M. tuberculosis. (A) Surface representation of LigD ligase domain (PDB 6NHZ) in the closed conformation. ATP (blue and orange sticks), magnesium (yellow spheres) and water molecules (red) are in the active site, surrounded by amino acids hydrogen bonding to the ATP. The active site (highlighted pink) with the catalytic lysine in magenta, is located within the NTase domain (grey) and capped by the OB domain (purple). (B) Surface representation of LigD ligase domain as in (A), with a nicked DNA substrate (purple). The NTase domain of human LigIV, bound to a DNA nick (PDB 6BKG) was aligned with the NTase domain of LigD (RMSD = 2.7 Å) to illustrate the possible location of the DNA nick near the active site (pink). (C) Structural alignment of the open conformation (PDB 1VS0, NTase dark blue, and OB purple) and closed conformation (PDB 6NHZ, NTase grey, and OB light purple) of the LigD ligase domain. Black arrow indicates rotation. Yellow sticks and spheres are ATP and magnesium from the closed conformation. (D) Active site of the LigD ligase closed conformation. The catalytic Lys481 has been mutated to methionine (magenta backbone). Amino acids interacting with the catalytic magnesium have a magenta backbone; amino acids that stabilize the active site through hydrogen bonding have a pink backbone. ATP, purple and orange sticks; magnesium, yellow spheres; water molecules, red spheres; black dashed lines indicate hydrogen bonding. (E) Active site of the LigD ligase domain in open conformation. The LIG-AMP intermediate has a magenta backbone, while amino acids interacting with the ribose sugar have a grey backbone and amino acids interacting with the adenine base have a pink backbone. Black dashed lines indicate hydrogen bonding Red sphere, water molecule. Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.

FIGURE 5

Atomic structure of the LigD polymerase domain. (A) Alignment of apo-LigD polymerase domain from M. tuberculosis (PDB 2IRU, purple) and P. aeruginosa (PDB 2FAO, grey), RMSD = 1.5 Å. (B) Electrostatic surface representation of M. tuberculosis LigD polymerase domain with dGTP and manganese in the active site (PDB 2IRY). Blue, positive charge; red, negative charge. (C) Active site of the polymerase domain from (B). Key amino acids of the active site are colored with a pink backbone. dGTP, magenta sticks; manganese, yellow spheres; water molecules, red spheres; black dashed lines indicate hydrogen bonding. Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.

FIGURE 6

Atomic structure of the LigD polymerase domain in a pre-catalytic state. (A) Cartoon representation of the M. tuberculosis polymerase domain (PDB 3PKY) with a 3′-overhang DNA substrate (magenta), UTP with a magenta backbone and manganese as yellow spheres. Key amino acids of the active site are shown with a pink backbone. (B) Overlay of the active sites from the polymerase domain in the presence of ATP (PDB 2IRY, grey) and after the LIG-AMP intermediate has formed (PDB 3PKY, purple). DNA and the AMP intermediate are in magenta. Key amino acids and manganese ions relevant to the pre-adenylylation complex are in grey. Amino acids key to the active site after LIG-AMP formation have a pink backbone, while manganese ions are yellow spheres. Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.

FIGURE 7

Atomic structures of the LigD polymerase domain from M. tuberculosis in complex with synapsed DNA substrates. (A) Surface representation of a chemically competent polymerase domain synaptic complex (PDB 4MKY). Each polymerase domain and associated dsDNA is individually colored. The pre-catalytic complex of the polymerase domain was aligned with the synaptic complex to indicate where the incoming nucleotide and metal ions would reside in relation to the DNA substrate (aligned with PDB 3PKY, RMSD = 0.37 Å). Key amino acids that stabilize the synaptic complex are highlighted in purple. (B) Surface representation of a chemically incompetent polymerase domain synaptic complex (PDB 2R9L). Each polymerase domain and associated dsDNA is individually colored. The pre-catalytic complex of the polymerase domain was aligned with the synaptic complex to indicate where the incoming nucleotide and metal ions would reside in relation to the DNA substrate (aligned with PDB 3PKY, RMSD = 0.34 Å). Key amino acids that stabilize the synaptic complex are highlighted in purple. Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.

FIGURE 8

Atomic structure of the LigD phosphoesterase domain from P. aeruginosa (PDB 3N9B). (A) Cartoon representation of the phosphoesterase domain with a sulfate ion (yellow sticks), manganese ion (yellow sphere), and water molecules (red spheres) in the active site. Amino acids critical to the active site are shown as sticks. Amino acids with a purple backbone interact with the manganese ion. Amino acids with a brown backbone interact with the sulfate ion and amino acids with a grey backbone stabilize the active site structure. Black dashed lines indicate hydrogen bonding. Pink dashed lines represent a disordered loop that was absent from the structure. (B) Electrostatic surface representation of the phosphoesterase domain. Blue, positive charge; red, negative charge. (C) Close-up of the active site of the phosphoesterase domain, colored as in (A). Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.

FIGURE 9

In silico predictions of LigD atomic structure (A) M. tuberculosis LigD, predicted by AlphaFold (Jumper et al., 2021), displayed as both cartoon and surface representations. POL–blue; PE–pink; LIG–purple (B) Electrostatic surface representation of (A), aligned with the NTase fold of the ligase domain from PDB 6NHZ, the pre-catalytic polymerase domain from PDB 3PKY, and the phosphoesterase domain from PDB 3N9B. Active sites of each enzymatic domain are encircled by blue dashes. Electropositive–blue; Electronegative–red (C) P. aeruginosa LigD, predicted by ColabFold (Mirdita et al., 2021), displayed as both cartoon and surface representations. POL–blue; PE–pink; LIG–purple (D) Surface representation of P. aeruginosa LigD model aligned with the open conformation of the ligase domain from M. tuberculosis shown as a cartoon (PDB 1VS0, RMSD = 1.1 Å). Black arrows represent locations of potential flexible loops (E) Surface representation of P. aeruginosa LigD model aligned with the catalytically competent polymerase domain synaptic complex from M. tuberculosis, shown as a cartoon (PDB 4MKY, RMSD = 1.4 Å). Black arrows represent locations of potential flexible loops (F) Electrostatic surface representation of (C), aligned with the NTase fold of the ligase domain from PDB 6NHZ, the pre-catalytic polymerase domain from PDB 3PKY and the phosphoesterase domain from PDB 3N9B. Active sites of each enzymatic domain are encircled by blue dashes. Electropositive–blue; Electronegative–red (G) B. subtilis LigD, predicted by ColabFold (Mirdita et al., 2021), displayed as both cartoon and surface representations. POL–blue; LIG–purple (H) Electrostatic surface representation of (G), aligned with the NTase fold of the ligase domain from PDB 6NHZ and the pre-catalytic polymerase domain from PDB 3PKY. Active sites of each enzymatic domain are encircled by blue dashes. Electropositive–blue; Electronegative–red. Figures generated with The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC.