Function of the ATR N-terminal domain revealed by an ATM/ATR chimera

Abstract

The ATM and ATR kinases function at the apex of checkpoint signaling pathways. These kinases share significant sequence similarity, phosphorylate many of the same substrates, and have overlapping roles in initiating cell cycle checkpoints. However, they sense DNA damage through distinct mechanisms. ATR primarily senses single stranded DNA (ssDNA) through its interaction with ATRIP, and ATM senses double strand breaks through its interaction with Nbs1. We determined that the N-terminus of ATR contains a domain that binds ATRIP. Attaching this domain to ATM allowed the fusion protein (ATM*) to bind ATRIP and associate with RPA-coated ssDNA. ATM* also gained the ability to localize efficiently to stalled replication forks as well as double strand breaks. Despite having normal kinase activity when tested in vitro and being phosphorylated on S1981 in vivo, ATM* is defective in checkpoint signaling and does not complement cellular deficiencies in either ATM or ATR. These data indicate that the N-terminus of ATR is sufficient to bind ATRIP and to promote localization to sites of replication stress.

Fig. 1

The N-terminus of ATR is sufficient to promote ATRIP and RPA-ssDNA binding properties onto ATM. (A) Schematic diagram of ATR, ATM, and ATM*. (B) 293T cells were transfected with HA-ATM, HA-ATR, HA-ATM* and Myc-ATRIP as indicated. Following lysis, HA-immunoprecipitates were separated by SDS-PAGE and blotted with Myc and HA antibodies. (C) Cell lysates containing ATR, ATM or ATM* were incubated with ssDNA bound to sepharose beads in the presence (+) or absence (−) of recombinant RPA. After washing, the bound proteins were eluted, separated by SDS-PAGE and immunoblotted. 10% of the lysate input (In) is included for comparison.

Fig. 2

ATM* localizes to sites of double strand breaks and replication stress. ATR^flox/−cells stably expressing (A) HA-ATR, (B) HA-ATM, or (C) HA-ATM* were treated with ionizing radiation (IR) or hydroxyurea (HU). 3.5 hours after treatment cells were stained with HA and ATRIP antibodies and appropriate fluorescent secondary antibodies. The percentage of cells with (D) HA or (E) ATRIP foci were scored. Error bars represent the standard deviation.

Fig. 3

ATM* is an active kinase in vitro and autoophosphorylates in cells. (A) ATM, ATM kinase-dead (KD) or ATM* were immunoprecipitated from transiently transfected cells and immune-complex kinase assays were performed with recombinant GST-BRCA11351-1552 substrate. (B) Flag-ATM* was immunoprecipitated from transfected cells following treatment with IR, HU, or UV. Immunoprecipitates were blotted with antibodies to Flag and ATM phospho-S1981. (C) ATR^flox/− cells or ATR^flox/− cells stably expressing ATM* were treated with IR, HU, or UV. Cell lysates were immunoblotted with ATM or ATM phopsho-S1981 antibodies. (D) A-T cells or A-T cells expressing ATM* were treated with IR, HU, or UV. Cell lysates were harvested, separated by SDS-PAGE and blotted with ATM or ATM P-S1981 antibodies.

Fig. 4

ATM* does not complement A-T cells. (A) Cell lysates from stable A-T cell lines expressing wild-type ATM, or ATM* were immuno-blotted with antibodies to ATM to demonstrate equal expression. (B) A-T, A-T/ATM or A-T/ATM* cell lines were treated with IR. Cell lysates were separated by SDS-PAGE and immunoblotted with the indicated antibodies. (C) A-T, A-T/ATM or A-T/ATM* cell lines were treated with 0, 2, or 4Gy of IR and colony formation was assessed after 3 weeks. Error bars represent standard deviations. (D) A-T or A-T/ATM* cell lines were treated with 6 Gy of IR and their ability to arrest at the G2/M checkpoint was assessed. The percentage of cells in mitosis with and without ionizing radiation is shown. (E) 293T cells were transfected with HA-ATM*, HA-ATM, HA-ATR, and Myc-Nbs1 expression vectors as indicated. HA-immunoprecipitates were separated by SDS-PAGE and blotted with Myc and HA antibodies.

Fig. 5

ATM* does not complement ATR-deficient cells. (A) ATR^+/+, ATR^flox/−, or ATR^flox/−cells expressing ATM* or ATR were infected with adenovirus expressing either GFP as a control or CRE to delete the “floxed” ATR allele. Four days after infection cells were treated with 50J/m² of UV, incubated for 1 hour, and harvested. Following SDS-PAGE separation, cell lysates were immunoblotted with the indicated antibodies. (B) ATR^flox/−, or ATR^flox/− cells expressing ATM* were infected with GFP or CRE adenovirus then treated with HU, UV, or IR. Cell lysates were separated by SDS-PAGE and blotted with the indicated antibodies.

Fig. 6

ATM* is not activated by TopBP1. HA-ATR-ATRIP or HA-ATM*-ATRIP complexes were purified from cells and placed in immune-complex kinase reactions with GST-MCM2 as a substrate. Recombinant TopBP1 containing the ATR-activation domain and BRCT repeats 7–8 (TopBP1 AAD) or just BRCT repeats 7–8 (TopBP1 7&8) was added to the reaction as indicated. Following separation by SDS-PAGE the incorporation of ³²P into MCM2 was measured by autoradiography. Equal levels of MCM2 substrate was confirmed by coomassie staining and levels of HA-ATM or HA-ATM* were analyzed by immunoblotting with HA antibody.