Bcl2 induces DNA replication stress by inhibiting ribonucleotide reductase

Abstract

DNA replication stress is an inefficient DNA synthesis process that leads replication forks to progress slowly or stall. Two main factors that cause replication stress are alterations in pools of deoxyribonucleotide (dNTP) precursors required for DNA synthesis and changes in the activity of proteins required for synthesis of dNTPs. Ribonucleotide reductase (RNR), containing regulatory hRRM1 and catalytic hRRM2 subunits, is the enzyme that catalyzes the conversion of ribonucleoside diphosphates (NDP) to deoxyribonucleoside diphosphates (dNDP) and thereby provides dNTP precursors needed for the synthesis of DNA. Here, we demonstrate that either endogenous or exogenous expression of Bcl2 results in decreases in RNR activity and intracellular dNTP, retardation of DNA replication fork progression, and increased rate of fork asymmetry leading to DNA replication stress. Bcl2 colocalizes with hRRM1 and hRRM2 in the cytoplasm and directly interacts via its BH4 domain with hRRM2 but not hRRM1. Removal of the BH4 domain of Bcl2 abrogates its inhibitory effects on RNR activity, dNTP pool level, and DNA replication. Intriguingly, Bcl2 directly inhibits RNR activity by disrupting the functional hRRM1/hRRM2 complex via its BH4 domain. Our findings argue that Bcl2 reduces intracellular dNTPs by inhibiting ribonucleotide reductase activity, thereby providing insight into how Bcl2 triggers DNA replication stress.

Figure 1

Expression of endogenous Bcl2 is correlated with decreased levels of RNR activity and intracellular dNTPs. A, expression levels of Bcl2, hRRM1 and hRRM2 were analyzed by Western blot. B-C, extracts from various human lung cancer cells were incubated with ¹⁴C-CDP. The generation of ¹⁴C-dCDP was analyzed by a phosphorImager on TLC. The relative abundance of ¹⁴C-dCDP and ¹⁴C-CDP was quantified with imageQuant software. RNR activity was calculated by a formula as ¹⁴C-dCDP/(¹⁴C-CDP+¹⁴C-dCDP). Error bars represent ±SD of three repeated determinations. D, intracellular levels of dNTPs in various human lung cancer cells were measured. Error bars represent ± SD of three repeated determinations.

Figure 2

Bcl2 suppresses DNA replication fork progression. A, Bcl2 was stably expressed in H1299 human lung cancer cells. B, schematic diagram for calculation of the fork rate (Top) and inter origin distance (Bottom) of replication fork progression. C, H1299 cells expressing Bcl2 or vector-only control were pulsed-labeled with 100 μM CldU for 20 min and 100 μM IdU for another 20 min. The labeled cells were processed for DNA combing. Representative pairs of sister replication forks are shown. Red, CldU. Green, IdU. Bar, 5 μm (1 μm = 2.59 kb). D, distribution of fork rate (kb/min) in H1299 cells expressing Bcl2 or vector-only control. The mean±SD for fork rate and number of scores are summarized. The p value was determined by a two-tailed Mann-Whitney test. E, distribution of inter origin distances (kb) were compared in H1299 cells expressing Bcl2 or vector-only control. The mean ±SD for distances and number of scores are summarized. F, representative images of symmetrical and asymmetrical replication. G, distribution of the degree of asymmetry of bidirectional replication forks was compared in H1299 cells expressing Bcl2 or vector-only control.

Figure 3

Quantum dot-based immunofluorescence (QD-IF) analysis of co-localization of Bcl2 and RNR. Co-localization of Bcl2/hRRM1 (A) and Bcl2/hRRM2 (B) was measured by QD-IF in H460 cells expressing high levels of endogenous Bcl2, hRRM1 and hRRM2 as described in “Methods”. Quantification of co-localization was analyzed with Nuance imaging software (Caliper/PerkinElmer).

Figure 4

Bcl2 directly interacts with hRRM2 via the BH4 domain. A, purified WT Bcl2 (20 ng/ml) was incubated with purified hRRM1 or hRRM2 (20 ng/ml) in 1% CHAPS lysis buffer at 4°C for 2h. Co-immunoprecipitation (co-IP) was performed using Bcl2, hRRM1 or hRRM2 antibody, respectively. Bcl2, hRRM1 and hRRM2 were analyzed by Western blot. B-C, purified WT or each of the BH deletion Bcl2 mutant proteins (20 ng/ml) was incubated with 20 ng/ml of purified hRRM2 (B) or hRRM1 (C) in 1% CHAPS lysis buffer at 4°C for 2h. Co-IP was performed using agarose-conjugated Bcl2 antibody. Bcl2, hRRM1 and hRRM2 were analyzed by Western blot.

Figure 5

Bcl2 directly disrupts the hRRM1/hRRM2 complex, which requires the BH4 domain. A-B, the hRRM1/hRRM2 complex was co-immunoprecipitated using hRRM1 or hRRM2 antibody, respectively, from H1299 vector-only cells, and incubated with increasing concentrations of purified WT Bcl2 in 1% CHAPS lysis buffer at 4°C for 2h. After centrifugation, the resulting supernatant or immunocomplexed beads were subjected to SDS-PAGE. Bcl2, hRRM1 and hRRM2 on beads or in supernatant were analyzed by Western blot. C-D, the hRRM1/hRRM2 complex was obtained as above and incubated with purified WT or each of the BH deletion Bcl2 mutant proteins in 1% CHAPS lysis buffer at 4°C for 2h. After centrifugation and washing, the immunocomplexes on beads were subjected to SDS-PAGE and analyzed by Western blot to measure hRRM1-associated hRRM2 and hRRM2-associated hRRM1.

Figure 6

The BH4 domain is essential for Bcl2 to inhibit RNR activity, dNTPs synthesis and DNA replication fork progression. A, WT or each of the BH deletion Bcl2 mutants was stably expressed in H1299 cells. Expression of Bcl2, hRRM1 and hRRM2 were analyzed by Western blot. B-D, RNR activity and dNTP pools in H1299 cells expressing WT or each of the BH deletion Bcl2 mutants were measured and quantified as described in the legend of figure 1. Error bars represent ± SD of three repeated determinations. E, representative pairs of sister replication forks from H1299 cells expressing WT or each of the BH deletion Bcl2 mutants are shown. F-G, distributions of fork rate (kb/min) or inter origin distance in H1299 cells expressing WT or each of the BH deletion Bcl2 mutants were analyzed as described in the legend of figure 2.

Figure 7

Knockdown of endogenous Bcl2 upregulates RNR activity and intracellular levels of dNTPs, and accelerates DNA replication fork progression. A, Bcl2 shRNA or control shRNA was stably transfected in H460 cells. Expression levels of Bcl2, hRRM1and hRRM2 were analyzed by Western blot. B, H460 cells expressing Bcl2 shRNA or control shRNA were disrupted in 1% CHAPS lysis buffer. Co-IP was performed using hRRM1 or hRRM2 antibody, respectively. The hRRM1-associated hRRM2 and hRRM2-associated hRRM1 were analyzed by Western blot. C-E, RNR activity and dNTP pool in H460 cells expressing Bcl2 shRNA or control shRNA were analyzed as described in the legend of Figure 1. Error bars represent ±SD of three repeated determinations. F, representative pairs of sister replication forks from H460 cells expressing Bcl2 shRNA or control shRNA are shown. G-I, distributions of fork rate (kb/min), inter origin distance or asymmetry in H460 cells expressing Bcl2 shRNA or control shRNA were analyzed as described in the legend of figure 2.