Figure 1.
Identification of the ARF-interacting domain on topo I and the role of PS506 in ARF binding.
(A) Scheme showing regions of the topo I protein encoded by the 9 expression constructs. The N-terminal domain, core domain, linker, and C-terminal domain are indicated. (B) Analysis of expression (top), PS506 content (middle), and ARF binding (bottom) of the products of constructs 1–9 expressed in human H358 lung cancer cells. Cells were treated with Adp14 (20 moi) to elevate cellular ARF levels. Lysates were collected 2 days after treatment and subjected to mouse anti-FLAG immunoprecipitation (IP) followed by Western (W) analysis with antibodies specific for FLAG (rabbit anti-FLAG), PS506, or ARF, as indicated.
Figure 2.
Effects of topo I phosphorylation status and ARF on topo I DNA association and relaxation activity.
(A) Time course of non-covalent association [low salt (75 mM NaCl) at 4°C] of various R-topo I forms (0.3 pmol) with 0.03 pmol [3H]-labeled plasmid DNA in the presence or absence of 0.3 pmol bacterially produced human ARF fusion protein: hyperphosphorylated topo I minus ARF (▪) or plus ARF (□); basal phosphorylated topo I minus ARF (▾) or plus ARF (▿); and unphosphorylated (AP-treated) topo I minus ARF (•) or plus ARF (○). (B) Row 1: Topo I immunoprecipitation (IP) with basal phosphorylated (lane “B”), hyperphosphorylated (lane “CK2”), and unphosphorylated (lane “AP”) R-topo I (1 µg) followed by phosphoserine Western; row 2: topo I IP of basal phosphorylated, hyperphosphorylated, or unphosphorylated R-topo I (1 µg) incubated with 0.14 µg ARF, followed by ARF Western; rows 3 and 4: Western analyses of PS506 and total topo I, respectively, with the same basal phosphorylated, hyperphosphorylated, and unphosphorylated R-topo I samples as in rows 1 and 2 (0.3 µg per lane) (C) Rate of topo I-catalyzed nicking of a radiolabeled suicide substrate that traps topo I and DNA in a covalent complex. Non-covalent complexes of DNA with basal or hyperphosphorylated R-topo I were preformed by incubation in low salt at 4°C in the presence or absence of recombinant ARF, then the temperature was raised to 8°C for the indicated times. Covalently linked DNA-topo I complexes were recovered by precipitation with K+SDS and quantified by scintillation counting. (D) Topo I-mediated plasmid relaxation assay performed with unphosphorylated, basal phosphorylated, and hyperphosphorylated R- topo I, followed by agarose gel electrophoresis to separate substrate and products; s = supercoiled, r = relaxed plasmid DNA.
Figure 3.
Chromatin association of topo I.
(A) Rows 1–3: Topo I IP followed by Western analyses of total topo I, phosphoserine, and ARF was performed before (lane “C”) or 2 days after (lane “TBB”) treatment of H358 cells with TBB (10 µM for 1 h); rows 4 and 5: Western analyses of PS506 and total topo I in the same starting samples as in rows 1–3. Quantification of band densities indicated that TBB treatment reduced both P-ser and PS506 reactivity by ∼80%. (B) Histone H3 chromatin immunoprecipitation (ChIP) of untreated H358 cells (lane 1) or 2 days after treatment with 20 moi Adp14 (lane 2), TBB (10 µM, 1 h; lane 3), or both Adp14 and TBB (lane 4), followed by Western analyses of histone H3 and topo I. (C) The results of four independent ChIP analyses performed as in (B); bars represent the mean and standard deviation of chromatin-associated topo I levels in the treated cells relative to the untreated cells, quantified digitally from band intensities.
Figure 4.
Topo I cleavage complex formation and induction of DNA double-strand breaks.
(A) Cleavage complex formation in untreated OVCAR-3 cells (bars 1,5), or 2 days after treatment with 20 moi Adp14 (bars 2,6), 10 nM of the CK2 activator 1-ethyl, 4,5 dicarbamoyl imidazole (bars 3,7), or both Adp14 and the CK2 activator (bars 4,8). Samples 5–8 were also treated with 10 µM of the ROS inducer pyocyanin. Cells were pulsed with [3H]-thymidine to label DNA, cleavage complexes were captured by K+SDS precipitation, and DNA was quantified by scintillation counting. (B) Western analysis of γ-H2A.X, an indicator of DNA double-strand break formation, in lysates of H358 cells subjected to the same treatments 1–8 as in part (A). Total H2A.X levels are shown as a control.
Figure 5.
Model summarizing how oncogene-induced ARF expression and CK2-mediated topo I hyperphosphorylation can converge to enhance topo I–DNA association and topo I-facilitated DNA damage.
Cancer cells with elevated CK2 levels (CK2hi) accumulate a PS506-hyperphosphorylated form of topo I with increased DNA binding properties. Chronic oncogene activation in the absence of wild-type p53 leads to sustained elevation of ARF (ARFhi), which is unable to promote p53-mediated apoptosis but is available to bind to PS506-hyperphosphorylated topo I, further promoting the association of topo I with DNA. The enhanced binding increases the potential for topo I-facilitated DNA double-strand break (DSB) formation in the presence of elevated levels of reactive oxygen species (ROS) that accompany oncogene activation.