Recognition of DNA Termini by the C-Terminal Region of the Ku80 and the DNA-Dependent Protein Kinase Catalytic Subunit

Abstract

DNA double strand breaks (DSBs) can be generated by endogenous cellular processes or exogenous agents in mammalian cells. These breaks are highly variable with respect to DNA sequence and structure and all are recognized in some context by the DNA-dependent protein kinase (DNA-PK). DNA-PK is a critical component necessary for the recognition and repair of DSBs via non-homologous end joining (NHEJ). Previously studies have shown that DNA-PK responds differentially to variations in DSB structure, but how DNA-PK senses differences in DNA substrate sequence and structure is unknown. Here we explore the enzymatic mechanisms by which DNA-PK is activated by various DNA substrates and provide evidence that the DNA-PK is differentially activated by DNA structural variations as a function of the C-terminal region of Ku80. Discrimination based on terminal DNA sequence variations, on the other hand, is independent of the Ku80 C-terminal interactions and likely results exclusively from DNA-dependent protein kinase catalytic subunit interactions with the DNA. We also show that sequence differences in DNA termini can drastically influence DNA repair through altered DNA-PK activation. These results indicate that even subtle differences in DNA substrates influence DNA-PK activation and ultimately the efficiency of DSB repair.

Fig 1. Ku80 Protein Structure, Mutant Construction, and Protein Purification.

a) Linear representation of Ku80 structural domains and mutant construction. b) Coomassie stained SDS gel of purified Ku and DNA-PKcs proteins used in this study. All Ku80 mutants were purified with wild type Ku70 which remains consistent in lane 1–4. Ku80 mutant constructs are indicated along the top of the gel and the Ku80 protein bands are denoted with red asterisks. c) Model of wild type and Ku mutant constructs used in this study. Ku80 is depicted in dark blue and Ku70 is depicted in grey. The structural regions of the Ku80 C-terminus are depicted in their corresponding color from (a).

Fig 2. Different Structural Regions of the Ku80 C-terminus Stimulate DNA-PKcs Activity Depending on the Structure of the DNA Substrate.

DNA-PK kinase activity is reported as pmol of phosphate transferred using the indicated DNA substrates. Reactions were completed with 30bp substrates containing a blunt (a) 3’ overhang (b) or 5’ overhang (c). DNA-PK activity was assessed on 60 bp substrates with blunt (d) 3’ or 5’ (e and f respectively. Y-substrates of 30 and 60bp are presented in panels g and h respectively. Ku constructs used in the reactions are indicated along bottom of the figure. Wild type bars are white, CLH bars contain horizontal lines, CL bars contain crosshatching, Core bars contain diagonal lines and no added Ku (-) bars are black. The 5’ and 3’ overhangs were prepared by digestion of the Blunt-ended gBlock with EcoRI and KpnI respectively. Sequences and other details concerning DNA substrates can be found in S1, S2, and S3 Tables. Data are presented as the mean and SD with asterisks indicating statistically significant differences compared to wild type (p <0.05).

Fig 3. DNA-PK activity stimulated by 400bp substrates.

DNA-PK kinase activity is reported as pmol of phosphate transferred using the indicated DNA substrates. Reactions were completed with 400bp substrates containing a blunt (a) 3’ overhang (b) or 5’ overhang (c). Sequences and other details concerning DNA substrates can be found in S1, S2, and S3 Tables. Data are presented as the mean and SD with asterisks indicating statistically significant differences compared to wild type (p <0.05).

Fig 4. The Ku80 C-terminus is Dispensable for DNA-PKcs-DNA Binding.

A modified ELISA was used to monitor DNA-PKcs-DNA binding and was performed as described in Materials and Methods. (a) Recruitment on 30bp DNA (b) Recruitment on a 60bp DNA. (c) DNA substrate influence on DNA-PKcs-DNA binding was tested on indicated substrates. Data are reported as the mean and SD of the rate of change in absorbance at OD_370nm in 15 minutes. (a-b) Asterisks indicate statistically significant differences compared to wild type (p<0.05). (c) Differences between 60bp vs 30bp substrate were statistically significant (p<0.05). (d), DNA binding Activity of Ku on 30bp and 60bp DNA. EMSA analysis of DNA binding activity was performed as described in Materials and methods. Ku was titrated at equal concentrations with the 30bp DNA (Lanes 1–4 and 60bp DNA (lanes 5–8). Products separated via native electrophoresis and imaged on a PhosphorImager. The free DNA is indicated by the bracket, and protein-DNA complexes representing a single Ku bound to a DNA and 2 Ku molecules bound to a DNA are indicated by the arrows.

Fig 5. Distinct Influences of the Ku80 C-terminus on DNA-PK Activation with Linearized Plasmid DNA.

a) DNA-PK kinase stimulation with plasmid DNA linearized with EcoRV generating blunt-ended termini and with KpnI generating 4 base 3’ single stranded overhangs. b) DNA-PK kinase stimulation with plasmid DNA linearized with XhoI and BamHI generating 4 base 5’ single stranded overhangs. DNA substrates are depicted pictorially. DNA termini generated by digestion are depicted below each graph indicating locations of pyrimidines (Py) and purines (Pu). Kinase activity is reported as the mean and SD of pmol of phosphate transferred. Asterisks indicate statistically significant differences compared to wild type (p <0.05). (b) Asterisk comparing wild type of XhoI digested DNA and wild type of BamHI digested DNA indicates statistically significant differences (p<0.05).

Fig 6. Preferential DNA-PK Activation by Terminal Pyrimidines Leads to Increased NHEJ.

a) DNA-PK kinase activity is determined with wild type Ku and linearized reporter plasmid digested with restriction enzymes XbaI and EcoRI. DNA termini generated by digestion are depicted below the graph indicating locations of pyrimidines (Py) and purines (Pu). Activity is reported as pmol of phosphate transferred. b) Representative images analyzed for host cell reactivation assay showing GFP and RFP expression. Images were obtained using a 40x objective. Scale bar = 50 μm. c) Quantified results from host cell reactivation assay. Results are reported as ratio of green cells to red fluorescent cells relative to NHEJ activity. Results from XbaI digested DNA are shown in blue and results from EcoRI digested DNA are shown in purple. Data is presented as the mean and SD with asterisks indicating statistically significant differences.

Fig 7. Model of Distinct Mechanisms of DNA-PK Activation Dependent on DNA Cofactor Structure.

1. The Ku heterodimer binds to DNA termini and undergoes a conformational change involving the C-terminus of Ku80 [14]. 2. DNA-PKcs is recruited to the DNA terminus through interactions with Ku but independent of interactions with the Ku80 C-terminus (See Fig 3). 3. When bound to blunt-ended DNA, DNA-PKcs is activated through interactions with the disordered linker, helical bundle, and extreme C-terminus of the Ku80 C-terminus. When bound to DNA with 3’ or 5’ overhangs, DNA-PKcs is activated through interactions with the disordered linker and helical bundle of the Ku80 C-terminus. The 5’ end of DNA activates the kinase while the 3’ end is free to anneal across the synapse [2]. A DSB with a 5’ overhang is depicted. DNA sequence specific interactions between the 5’ overhang and DNA-PKcs are indicated in the model.