Proteasome inhibition suppresses DNA-dependent protein kinase activation caused by camptothecin

Abstract

The ubiquitin-proteasome pathway plays an important role in DNA damage signaling and repair by facilitating the recruitment and activation of DNA repair factors and signaling proteins at sites of damaged chromatin. Proteasome activity is generally not thought to be required for activation of apical signaling kinases including the PI3K-related kinases (PIKKs) ATM, ATR, and DNA-PK that orchestrate downstream signaling cascades in response to diverse genotoxic stimuli. In a previous work, we showed that inhibition of the proteasome by MG-132 suppressed 53BP1 (p53 binding protein1) phosphorylation as well as RPA2 (replication protein A2) phosphorylation in response to the topoisomerase I (TopI) poison camptothecin (CPT). To address the mechanism of proteasome-dependent RPA2 phosphorylation, we investigated the effects of proteasome inhibitors on the upstream PIKKs. MG-132 sharply suppressed CPT-induced DNA-PKcs autophosphorylation, a marker of the activation, whereas the phosphorylation of ATM and ATR substrates was only slightly suppressed by MG-132, suggesting that DNA-PK among the PIKKs is specifically regulated by the proteasome in response to CPT. On the other hand, MG-132 did not suppress DNA-PK activation in response to UV or IR. MG-132 blocked the interaction between DNA-PKcs and Ku heterodimer enhanced by CPT, and hydroxyurea pre-treatment completely abolished CPT-induced DNA-PKcs autophosphorylation, indicating a requirement for ongoing DNA replication. CPT-induced TopI degradation occurred independent of DNA-PK activation, suggesting that DNA-PK activation does not require degradation of trapped TopI complexes. The combined results suggest that CPT-dependent replication fork collapse activates DNA-PK signaling through a proteasome dependent, TopI degradation-independent pathway. The implications of DNA-PK activation in the context of TopI poison-based therapies are discussed.

Fig 1

Suppression of RPA2 hyperphosphorylation by MG-132. (A) The N-terminus sequence of RPA2. Eight serines and a threonine are the target of PIKKs and CDK. (B) MG-132 suppressed CPT-induced hyperphosphorylation of RPA2. RPA2 hyperphosphorylation was analyzed by Western blotting in CPT (2 μM, 1 h)-treated HeLa cells with or without MG-132 (25 μM, 1 h) pretreatment. (C) Bypass of CDK-mediated phosphorylation does not reverse MG-132 suppression of RPA2 phosphorylation. RPA2 proteins mutated on CDK phosphorylation sites (S23/29A and S23/29D) were expressed in 293T cells and analyzed phosphorylation by Western blotting using Flag antibody after CPT (2 μM, 1 h) treatment. (D, E) MG-132 does not affect RPA foci formation. HeLa cells were treated with CPT (2 μM, 1 h) after mock or MG-132 (25 μM, 1 h) treatment and stained with RPA2 antibody and phospho-specific antibody against Ser4/8 of RPA2 (D). Graphical data of the percentage of cells with RPA2 foci (E). Error bars show S.E. calculated from three independent experiments.

Fig 2

CPT-induced DNA-PK activation is abrogated by proteasome inhibitors. (A) Suppression of DNA-PK activation by MG-132. HeLa cells were pretreated with vehicle or MG-132 (25 μM, 1 h) and CPT (2 μM, 1 h)-induced DNA-PKcs, ATM, and Chk1 phosphorylation were analyzed by Western blotting using the indicated antibodies. (B) DNA-PKcs Ser2056 is an autophosphorylation site in response to CPT. HeLa cells were pretreated with DNA-PK inhibitor (NU7026, 10 μM) or ATM inhibitor (KU55399, 10 μM) for 1 h followed by CPT (2 μM, 1 h) treatment. DNA-PKcs and ATM phosphorylation were analyzed by Western blotting. (C) Suppression of DNA-PK by ALLN and Bortezomib. DNA-PK and RPA2 phosphorylation were analyzed by Western blotting in cells treated with MG-132 (25 μM, 1 h), ALLN (50 μM, 1 h), or Bortezomib (100 nM, 2 h). (D) MG-132 does not affect UV- or IR-induced DNA-PK activation. DNA-PKcs autophosphorylation was analyzed by Western blotting using extracts prepared from cells UV irradiated (20 J/m²) or exposed to 5 Gy IR. (E) MG-132 blocks CPT-induced association of DNA-PKcs and Ku70. HeLa cells were treated with CPT (2 μM, 1 h) and whole cell extracts were prepared. DNA-PKcs was immunoprecipitated and coimmunoprecipitated Ku70 was detected by Western blotting.

Fig 3

DNA-PK activation is dependent on DNA replication, but not S phase specific regulation. (A) DNA replication-dependent activation of DNA-PK. CPT (2 μM), HU (4 mM), and MG-132 (10 μM) were added to the media of HeLa cells in the indicated combinations. DNA-PKcs and RPA2 phosphorylation and Top1 protein were analyzed by Western blotting. (B) DNA replication in the presence of MG-132. BrdU-incorporated cells were stained with specific antibody after 1 h treatment of MG-132 (25 μM). (C) Replication-sensitive and MG-132-resistant γH2AX induced by CPT. Cells were treated with CPT (2 μM, 1 h) following mock, MG-132 (10 μM, 1 h), or HU (4 mM, 15 min) pretreatment and stained with γH2AX antibody. The percentage of cells with γH2AX was averaged from three independent experiments. Error bars are the S.E. (D) MG-132 does not affect NCS-induced DNA-PK activation in S phase. U2OS cells were released from thymidine block to synchronize in S phase and analyzed for DNA-PKcs autophosphorylation induced by CPT (2 μM, 1 h) or NCS (100 ng/ml, 1 h). Thy− and Thy+ indicate cells without and with thymidine treatment, respectively, to synchronize in S phase.

Fig 4

DNA-PK and 53BP1 are independently regulated by the proteasome. (A) 53BP1-independent DNA-PK activation in response to CPT. HeLa cells were transfected with control or 53BP1 siRNA, and CPT (2 μM, 1 h)-induced DNA-PKcs autophosphorylation was analyzed by Western blotting. (B) DNA-PK-independent foci formation of 53BP1. HeLa cells were treated with vehicle or the DNA-PK inhibitor NU7026 (10 μM, 1 h) prior to treatment with CPT (2 μM, 1 h). 53BP1 foci formation was examined by immunofluorescence microscopy.