Ancient and recent adaptive evolution of primate non-homologous end joining genes

Abstract

In human cells, DNA double-strand breaks are repaired primarily by the non-homologous end joining (NHEJ) pathway. Given their critical nature, we expected NHEJ proteins to be evolutionarily conserved, with relatively little sequence change over time. Here, we report that while critical domains of these proteins are conserved as expected, the sequence of NHEJ proteins has also been shaped by recurrent positive selection, leading to rapid sequence evolution in other protein domains. In order to characterize the molecular evolution of the human NHEJ pathway, we generated large simian primate sequence datasets for NHEJ genes. Codon-based models of gene evolution yielded statistical support for the recurrent positive selection of five NHEJ genes during primate evolution: XRCC4, NBS1, Artemis, POLλ, and CtIP. Analysis of human polymorphism data using the composite of multiple signals (CMS) test revealed that XRCC4 has also been subjected to positive selection in modern humans. Crystal structures are available for XRCC4, Nbs1, and Polλ; and residues under positive selection fall exclusively on the surfaces of these proteins. Despite the positive selection of such residues, biochemical experiments with variants of one positively selected site in Nbs1 confirm that functions necessary for DNA repair and checkpoint signaling have been conserved. However, many viruses interact with the proteins of the NHEJ pathway as part of their infectious lifecycle. We propose that an ongoing evolutionary arms race between viruses and NHEJ genes may be driving the surprisingly rapid evolution of these critical genes.

Figure 1. Sliding window analysis identifies five candidate NHEJ genes evolving under positive selection.

A) A schematic of the mammalian non-homologous end joining pathway is shown, illustrating the roles of all proteins included in this study. B) A cladogram shows the relationship of the primate species used in the sliding window analysis. Branch colors correspond to the sliding window comparisons graphed in panel C. C) Sliding window analysis of dN/dS along the length of KU70 and XRCC4. In each case, three pairwise sequence alignments were analyzed (human - orangutan comparison in pink, human - rhesus comparison in green, and rhesus - marmoset comparison in orange). D) The table summarizes the maximum dN/dS peak height found along the length of each pairwise sliding window comparison made. Asterisks indicate statistically significant peaks (p<0.05). In gray highlight are the five genes with significant peaks in at least two out of three comparisons.

Figure 2. Five proteins in the NHEJ pathway show signatures of positive selection.

Domain diagrams are shown for the five NHEJ proteins evolving under recurrent positive selection during primate speciation. The locations of specific amino acid positions under positive selection have been marked on these diagrams (red tick mark indicates posterior probability of >0.95, black tick mark indicates posterior probability of >0.90). Amino acid positions specifically discussed in the text are indicated.

Figure 3. Residues under positive selection in Polλ fall on the protein surface.

The co-crystal of the human Polλ 39 kDa catalytic domain in complex with a DNA substrate is shown (PDB 2BCQ) . Four of the eight amino acid positions found to be under positive selection (330, 381, 441, and 484; red globes) could be mapped onto the structure. All four fall onto the outer surface of the protein and are not predicted to interfere with the polymerase active site.

Figure 4. XRCC4, a component of the NHEJ ligase complex, shows a clustered signature of positive selection.

A) A co-crystal of the human XRCC4 homodimer (grey) in complex with a fragment of its binding partner Lig4 (blue) has been solved (PDB 1IK9) . The ligase-binding domain of XRCC4 is shown in yellow. The C-terminal domain of the 336 amino acid protein is unstructured and had to be truncated for crystallization. This portion has been artificially indicated by the wavy black line. In the crystal structure, the two monomeric chains are different lengths. Chain A (dark gray) is comprised of residues 1–211, while chain B (light gray) is comprised of residues 1–201. Two of the amino acids positions found to be under positive selection (205 and 211; red globes) could be mapped only to the longer of the two monomers (chain A). Sites 216, 218, 243, and 292 could not be mapped to either monomer. Their approximate location is marked with a pink asterisks on the linear schematic of the C-terminal domain. B) A linear domain diagram of XRCC4 is shown, with the approximate location of the amino acid sites under positive selection marked with asterisks. An amino acid alignment in this region for the 20 primate species used in this study is shown, with residues found to be under positive selection highlighted in gray. Residue 211, which was identified as being subject to positive selection, lies at the third position within the SUMOylation consensus site (IKQE; denoted in red), with the neighboring lysine being SUMOylated . Another amino acid position that has evolved under positive selection, residue 243, is located just four positions upstream of a A247S human disease mutation which has been linked to oral cancer susceptibility , and three positions downstream of the human Q240P polymorphism (these two sites are underlined in the human amino acid sequence). Secondary structure predictions and confidence values (0, low; 9,high) were obtained with the PSIPRED server . “H” and the barrel shape denote the very end of the long alpha helix that is observed in the crystal structure. Downstream of this, “C” indicates the unstructured region.

Figure 5. XRCC4 is under positive selection in modern humans.

CMS analysis of XRCC4 in the CEU population. The CEU population represents humans with ancestry from northern and western Europe. Bars on the x-axis indicate genes (red bar indicates XRCC4; grey bar indicates VCAN), and black dots show CMS values.

Figure 6. Interactions with other repair proteins have been conserved in Nbs1 despite its positive selection.

A) Positively selected residues 9 and 185 (red balls) are mapped onto the partial Nbs1 structure (PDB 3HUE) . B) SNP frequencies of Q185E are reported for the ten human populations included in the HapMap project (http://hapmap.ncbi.nlm.nih.gov/). Three-letter labels are standard codes (ASW - African ancestry in Southwest USA; CEU - Utah residents with Northern and Western European ancestry; CHB- Han Chinese in Beijing, China; CHD - Chinese in Metropolitan Denver, Colorado; GIH - Gujarati Indians in Houston, Texas; JPT - Japanese in Tokyo, Japan; LWK - Luhya in Webuye, Kenya; MEX - Mexican ancestry in Los Angeles, California; MKK - Maasai in Kinyawa, Kenya; TSI - Toscans in Italy). C) Binding assays were performed between recombinant biotinylated MRN complexes containing Nbs1 E185 or Q185, and an N-terminal Flag-tagged fragment of Mdc1 containing amino acids 1 to 740, as indicated. The biotinylated MRN complexes (20nM) were incubated with 45 nM Mdc1 and then isolated with streptavidin-coated magnetic beads. Bound protein was visualized by western blotting with anti-Flag (Mdc1) and anti-Nbs1 antibodies. D) MRN complexes containing Nbs1 E185 or Q185 were tested in ATM kinase assays with linear DNA as indicated. Phosphorylation of the substrate, GST-p53 (aa 1–100), was assessed by western blotting using a phospho-specific antibody directed against p53-phospho-ser15 as previously described .