Archives
Earlier studies showed that exposure of cells
Earlier studies showed that exposure of cells to IR caused ATM-dependent phosphorylation of 53BP1, as judged by electrophoretic mobility shift [24], [25], [26]. To date, the only known in vivo 53BP1 phosphorylation site(s) are Ser25 and possibly Ser29 [27]. In the course of our studies, we noticed that a mutant 53BP1 protein, in which Ser25 and Ser29 are mutated to alanine residues, is still hyperphosphorylated in response to DNA damage. Here we report phosphorylation of 53BP1 at several novel residues, using mass spectrometry and phospho-specific antibodies, and show that ionising radiation-stimulated phosphorylation of these residues requires ATM. Although it is thought to be specific for DSBs, 53BP1 was found to be efficiently phosphorylated at several novel sites in response to UV-irradiation in an ATM-independent, ATR-dependent manner.
Materials and methods
Results and discussion
We found that 53BP1 in which Ser25 and Ser29 are mutated to alanines is still phosphorylated after exposure of cells to IR (data not shown). To map new IR-induced 53BP1 phosphorylation sites, HA-tagged 53BP1 was expressed in HEK293 cells by transient transfection and immunoprecipitated with anti-HA madecassol from extracts of cells that were exposed to IR or not. Precipitates were subjected to SDS-PAGE (Fig. 1A) and 53BP1 was excised and digested with trypsin. Tryptic peptides were analysed on a 4000 Q-Trap mass spectrometer using precursor ion scanning to identify potential phosphopeptides that were then identified by ms/ms. This revealed eight basal sites of phosphorylation in 53BP1 and three sites whose phosphorylation increased after treatment of cells with IR (Fig. 1B and C). All of the IR-inducible sites, Thr302, Ser831 and Ser1219 conformed to the S/T–Q motif phosphorylated by ATM, ATR and DNA-PK (DNA-dependent protein kinase, a relative of ATM and ATR [28]). Intriguingly, the basal phosphorylation sites were mostly serine residues followed either by Q or P (Fig. 1C). Ser/Thr–Pro motifs are potential sites of phosphorylation by MAP kinase family members and cyclin-dependent kinases. The Ser/Thr–Pro sites we identified were found not to be regulated by DNA damage (data not shown); phospho-specific antibodies raised against these residues recognised 53BP1 in cell extracts but this signal did not change after exposure of cells to a variety of genotoxins (data not shown). Ser25, that was previously shown to be phosphorylated after DNA damage [27] did not emerge from our mass spectrometric analysis, probably because of the properties of the tryptic phosphopeptide bearing this residue (data not shown).
Alignment of 53BP1 from humans, mice and chickens showed that Thr302 and Ser1219 are conserved in all three species, whereas Ser831 is not. Interestingly, although there is not a high degree of sequence conservation outside the Tudor and BRCT domains of 53BP1, several small blocks of homology can be seen in this region and several of these contain S/T–Q motifs: Ser13, Ser25, Ser166, Ser176/178, Thr302, Ser452, Ser523, Thr543, Thr1171 and Ser1219 (Fig. 1D). Of these, Ser25 is the only previously reported site of phosphorylation on 53BP1 [27]. Conservation around these sites suggests that these regions are functionally important. To further investigate the IR-induced phosphorylation of 53BP1, phospho-specific antibodies were raised against Thr302, Ser831 from our mass spectrometric analysis, and against Ser166, a combination of Ser176/178 and Ser452 that lie in conserved patches in 53BP1. All antibodies were affinity purified using the phosphopeptide immunogen. As shown in Fig. 2A, all of the purified antibodies recognised the phosphopeptide immunogen but not the corresponding non-phosphopeptide in dot–blot analysis. Furthermore, these antibodies all recognised transiently transfected wild-type HA-53BP1 in extracts of cells treated with IR, but not when the relevant phosphorylated serine was mutated to alanine (Fig. 2B).
Having ascertained the specificity of the 53BP1 phospho-specific antibodies, phosphorylation of endogenous 53BP1 was examined. Cells were exposed to IR and allowed to recover for different times before cells were lysed and extracts subjected to SDS-PAGE followed by Western blotting. As shown in Fig. 3A, phosphorylation of 53BP1 at Thr302, Ser831, Ser166, Ser176/Ser178 and Ser452 was apparent 15min after exposure to IR and phosphorylation of these residues was still evident 2h (Fig. 3A) and 4h (data not shown) post-irradiation. The kinetics of 53BP1 phosphorylation was similar to those of IR-induced phosphorylation of p53 Ser15 and SMC1 Ser966 (Fig. 3A). Similar results were obtained in U2OS cells and in HCT116 cells (data not shown). Addition of protein phosphatase to cell extracts abolished recognition of 53BP1 by each antibody (data not shown).