The drinking water contaminant dibromoacetonitrile delays G1-S transition and suppresses Chk1 activation at broken replication forks

Abstract

Chlorination of drinking water protects humans from water-born pathogens, but it also produces low concentrations of dibromoacetonitrile (DBAN), a common disinfectant by-product found in many water supply systems. DBAN is not mutagenic but causes DNA breaks and elevates sister chromatid exchange in mammalian cells. The WHO issued guidelines for DBAN after it was linked with cancer of the liver and stomach in rodents. How this haloacetonitrile promotes malignant cell transformation is unknown. Using fission yeast as a model, we report here that DBAN delays G1-S transition. DBAN does not hinder ongoing DNA replication, but specifically blocks the serine 345 phosphorylation of the DNA damage checkpoint kinase Chk1 by Rad3 (ATR) at broken replication forks. DBAN is particularly damaging for cells with defects in the lagging-strand DNA polymerase delta. This sensitivity can be explained by the dependency of pol delta mutants on Chk1 activation for survival. We conclude that DBAN targets a process or protein that acts at the start of S phase and is required for Chk1 phosphorylation. Taken together, DBAN may precipitate cancer by perturbing S phase and by blocking the Chk1-dependent response to replication fork damage.

Figure 1

Dibromoacetonitrile (DBAN) arrests cell cycle progression in a concentration dependent manner. Wild type cells (ade6-M210 leu1–32 ura4-D18) were synchronised in G2 by lactose gradient centrifugation and released into rich medium (3% glucose, 0.5% yeast extract, 100 mg/L adenine) without (UT = untreated) or with 10 μM of the indicated haloacetonitriles (HANs). All HANs were diluted from a 12 mM stock solution in DMSO to a final concentration of 10 μM. (A) bromoacetonitrile (BAN), (B) chloroacetonitrile (CAN), (C) dichloroacetonitrile (DCAN), (D) dibromoacetonitrile (DBAN), (E) trichloroacetonitrile (TCAN), (F) DBAN at 10 μM or 20 μM. (G) DBAN affects cells in S phase which triggers a second cycle delay.

Figure 2

DBAN and TCAN delay G1-S transition. (A) Wild type cells without auxotrophic markers were synchronised in G1 by nitrogen starvation in minimal medium (3% glucose, 0.67% nitrogen base w/o amino acids and ammonium sulphate) at 30 °C. Cells were washed and released at T = 0 h into pre-warmed minimal medium with ammonium sulphate. Samples were withdrawn at the indicated time points and the DNA content was measured by flow cytometry. The dotted lines indicate 1 copy of the chromosomes (1 C, G1) and two copies (2 C, G2), respectively. (B,C) DNA content at the start of the experiment (0 h), at 4 h and 8 h post-release. Untreated cells (UT) progress from G1 (1 C) to G2 (2 C) within 4 h. This progression is slowed down by DBAN but blocked by TCAN (indicated by the arrows). The final concentration of all HANs is 10 μM. In a control experiment, cells were arrested in early S (1 C) with 15 mM hydroxyurea (HU).

Figure 3

DBAN kills cells mutated in DNA polymerase delta. (A) Model of the replication fork (adapted from). (B–E) Serial dilutions of the indicated strains were applied to rich medium plates at 30 °C. One plate was incubated at 37 °C. Incubation time: 3 days. The final concentration of DBAN was 10 μM, but 20 μM for TCAN. The mutant alleles are: pol alpha (swi7.H4), Ctf4 (mcl1-1), pol delta (cdc6.23, cdc27.P11, cdc1.p13), pol epsilon (cdc20.M10), S phase cell cycle inhibitor Mik1 (mik1::ura4+ ), cell cycle inhibitor Wee1 (wee1::ura4+ ), Δcdm1 (non-essential pol delta subunit, cdm1::ura4+ ), MCM2 (cdc19.P1), MCM4 (cdc21-M68), MCM5 (nda4-108), DDK/Cdc7 (hsk1-1312), Rad4 (TopBP1) (rad4.116), MAP kinase Sty1 (sty1::ura4+ ). Final concentration of BAN, CAN and DCAN is 10 μM.

Figure 4

DBAN advances DNA replication in a pol delta mutant. (A) Wild type cells (ade6-M210 leu1–32 ura4-D18) were grown in rich medium at 30 °C and synchronised in early S phase for 3.5 h in 15 mM hydroxyurea (HU). HU was washed out and cells were released into pre-warmed rich medium. The DNA content was measured at the indicated times. The dotted lines indicate 1 copy of the chromosomes (1 C, G1) and two copies (2 C, G2), respectively. The DNA content of nitrogen starved cells (no N) was measured in a parallel experiment to have an internal standard. (B–D). The indicated mutant strains were HU-synchronised and released into rich medium at 30 °C with or without 10 μM DBAN. The green histogram is the DNA content of untreated cells, the red histogram is the DNA content in the presence of DBAN. The brown colour indicates that both histograms overlap.

Figure 5

DBAN suppresses Chk1 phosphorylation. (A,B) cds1-His ₆ HA ₂ (55 kDa) and chk1-HA ₃ (60 kDa) cells were grown in rich medium at 30 °C and treated for 4 h with 10 μM DBAN, 12 μM camptothecin (CPT), 12 mM hydroxyurea (HU) or the combination as indicated. Total protein extracts were loaded on a 6% phostag gel. Full phostag gels are shown. H2AX-S129-P was detected on a 20% acrylamide (37.5:1 acrylamide:bisacrylamide) gel. The Chk1 shift was detected on a 10% (100:1 acrylamide:bisacrylamide) gel. H2AX and Chk1 (normal) panels were cropped. The full images are shown in Figs S3 and S4, respectively. The arrows indicate the smaller Cds1 band, Chk1-S345 phosphorylation and the hyper-phosphorylated forms of Chk1. (C) Model of Chk1 activation by Rad3 at broken replication forks. CPT immobilises Topoisomerase 1 and the Rad9-Rad1-Hus1 ring aids Chk1 phosphorylation. Rad9 is also phosphorylated by Rad3. (D) Drop test of the indicated strains. Chk1-D155E-HA₃ is a kinase-dead mutant. DBAN: 10 μM; CPT: 12 μM. (E,F) chk1-HA ₃ cells were HU synchronised (3.5 h, 15 mM HU, rich medium) and released into medium without (UT) or with CPT (12 μM), CPT (12 μM) + DBAN (10 μM) or CPT (12 μM) + TCAN (10 μM). Total protein extracts were separated on a 8% (Mrc1) and 10% (Chk1) acrylamide gel. P = Phospho-Chk1-S345 shift band). (G) rad9-HA ₃ cells were HU-synchronised and released into rich medium without (UT) or with CPT (12 μM), DBAN (10 μM) or CPT (12 μM) + DBAN (10 μM). P = Phospho-Rad9 shift band. Full images: Mrc1: Fig. S5; Chk1: Fig. S6; Rad9: Fig. S7.

Figure 6

(A) DBAN blocks G1-S transition. (B) DBAN prevents Chk1 phosphorylation by Rad3 at a broken fork. The details of the model are discussed in the main text.