Activation of ATR and related PIKKs

Abstract

The DNA damage response kinase ATR is an essential regulator of genome integrity. TopBP1 functions as a general activator of ATR. We have recently shown that TopBP1 activates ATR through its regulatory subunit ATRIP and a PIKK regulatory domain (PRD) located adjacent to its kinase domain. This mechanism of ATR activation is conserved in the S. cerevisiae ortholog Mec1. ATR is a member of the PIKK family of protein kinases that includes ATM, DNA-PKcs, mTOR, and SMG1. The PRD regulates the kinase activity of other PIKKs and may serve as a site of interaction between these kinase and their respective activators. Activation of ATR by TopBP1 is maximal at low substrate concentrations and declines exponentially as substrate concentration increases. These data are consistent with a model in which TopBP1 acts to alter the conformation of ATR-ATRIP to increase the ability of ATR to bind substrates. A further understanding of the mechanism of ATR activation will likely provide insights into the regulation of related PIK kinases.

Introduction

The DNA damage response pathway ensures the faithful maintenance and replication of the genome in spite of genotoxic stress. This signaling pathway coordinates cell cycle transitions, DNA replication, DNA repair, gene transcription, and, in some cases, promotes apoptosis and senescence to eliminate damaged cells. At the apex of this pathway are two closely related protein kinases, ATM (Ataxia-Telangiectasia Mutated) and ATR (ATM and Rad3-related). ATM is activated primarily in response to DNA double-strand breaks. On the other hand, ATR responds to many types of DNA damage and replication stress including breaks, crosslinks, and base adducts. ATR senses abnormally long-stretches of single-stranded DNA that arise from the functional uncoupling of helicase and polymerase activities at replication forks or from the processing of DNA lesions such as the resection of double-strand breaks.^{1, 2}

ATR forms a stable complex with ATRIP (ATR-interacting protein), which regulates the localization of ATR to sites of replication stress and DNA damage and is essential for ATR signaling.^{3, 4} Activation of ATR also requires the activator protein TopBP1, which has dual roles in the initiation of DNA replication and the DNA damage response. Localization of TopBP1 to sites of replication stress and DNA damage occurs independently of ATR through its interaction with the PCNA-like DNA clamp called the 9-1-1 complex.^{5, 6} TopBP1 directly stimulates the kinase activity of the ATR-ATRIP complex towards multiple substrates. A region of TopBP1 called the ATR Activation Domain (AAD) is sufficient to activate ATR.⁷

TopBP1 Activation of ATR-ATRIP

We have recently examined how TopBP1 regulates the ATR-ATRIP complex and found that it requires conserved domains in both ATRIP and ATR. ATRIP promotes the association of ATR and TopBP1. The TopBP1-interacting region on ATRIP is necessary for the activation of ATR in vitro and normal ATR-dependent cellular responses to replication stress and DNA damage, such as the G2/M checkpoint in cells.⁸ We also determined that a previously uncharacterized domain of ATR adjacent to its kinase domain named the PRD (PIKK Regulatory Domain) is necessary for TopBP1-mediated activation of ATR. Defined mutational analysis of the ATR PRD revealed that it is not necessary for the kinase activity of ATR, but is required for the activation of ATR by TopBP1. Mutation of a single residue in this domain abolished ATR signaling in response to replication stress and DNA damage. Furthermore, an intact ATR PRD is necessary for the viability of proliferating cells in the absence of any DNA damaging agents.⁸ These data suggest that TopBP1 activation of ATR may be necessary every S-phase perhaps to regulate the firing of replication origins or respond to DNA replication problems resulting from endogenous DNA damage.⁹ Overall, it seems that the TopBP1 activation-step is required for all known functions of the ATR-ATRIP complex.

TopBP1 clearly has additional roles in maintaining the genome that are independent of ATR, such as recruiting replication proteins to origins to promote replication initation.¹⁰ Whether the interaction between TopBP1 and ATR-ATRIP affects the role of TopBP1 in replication has not been explored. TopBP1 is also a substrate of both ATR and ATM. An ATR/ATM phosphorylation site in the TopBP1 AAD potentiates the ability of TopBP1 to activate ATR at least when ATR is activated in an ATM-dependent manner in response to double strand breaks.¹¹ Finally, TopBP1 contains multiple BRCT repeats, which usually function in tandem as phospho-protein interacting domains. The first pair of BRCT repeats interacts with the Rad9 subunit of the 9-1-1 complex.^{5, 6} The other BRCT repeats likely interact with as yet unidentified proteins. Discovering additional replication and/or genome maintenance proteins that interact with TopBP1 will be critical to understanding its role in multiple pathways.

Many components of the DNA damage response pathway are conserved in all eukaryotic cells. In S. cerevisiae, the orthologs of ATR and ATRIP are Mec1 and Ddc2. However, it has not been clear whether the TopBP1 homolog Dpb11, which also functions in DNA replication and the DNA damage response, can serve as a Mec1 activator.¹² Dpb11 has not been shown to interact with Mec1 or Ddc2 and does not contain any obvious homology to the TopBP1 AAD. Our results indicate that the TopBP1 interacting region of ATRIP is functionally conserved in Ddc2.⁸ Furthermore, we have observed that Dpb11 can stimulate the kinase activity of Mec1 (D.M., unpublished data). Thus, the TopBP1/Dpb11 mechanism of ATR/Mec1 activation is conserved throughout evolution. In budding yeast, a subunit of the 9-1-1 clamp, Ddc1, can also function as a Mec1 activator.¹³ Whether the 9-1-1 complex directly activates ATR in other organisms remains to be determined. It will be interesting to see if the mechanism of Ddc1 activation is similar or distinct from that of Dpb11 and if it requires the PRD of Mec1.

A Common Mechanism of PIK Kinase Regulation

ATR is a member of the PIKK (phosphoinositide 3-kinase related kinase) family of protein kinases. The PIKK family also includes ATM, DNA-PKcs, mTOR, and SMG1, and orthologs of these proteins are often highly conserved. PIKKs are key regulators of a number of diverse cellular processes. As mentioned above, ATR and ATM coordinate the DNA damage response. DNA-PKcs (DNA-dependent protein kinase catalytic subunit) is also involved in maintaining genomic stability through its role in promoting DNA double-strand break repair by nonhomologous end-joining. mTOR (mammalian target of rapamycin) responds to amino acids levels and mitogenic stimuli to promote cell growth. Finally, SMG1 is a central component of nonsense-mediated mRNA decay, a conserved mRNA surveillance mechanism.

As the name implies, these atypical kinases contain a conserved catalytic domain, which has sequence similarity to the kinase domain of PI3 (phosphoinositide 3) lipid kinases. With the exception of mTOR, they preferentially target substrates at serines or threonines followed by a glutamine residue (SQ/TQ sites). The PIKKs also exhibit sequence homology in two regions that flank the kinase domain called the FAT (FRAP, ATM, TRRAP) and FATC (FAT C-terminus) of unknown function. The FATC domain is required for the kinase activity of all PIKKs.^{8, 14-17} One possibility is that it may interact with the FAT domain to promote the proper folding of the kinase domain. It could also be important for binding substrates.

Many studies have explored how these kinases are regulated, and a common theme is emerging. PIKKs are generally regulated by changes in their subcellular localization and an interaction with an activator protein or nucleic-acid-protein complex. ATR, ATM, and DNA-PKcs are recruited to sites of DNA damage through interactions with their respective interacting partners, ATRIP, NBS1, and Ku80.¹⁸ Recently, Rag GTPases were shown to promote the localization of mTOR to distinct perinuclear compartments.¹⁹ Activation of each PIKK requires a unique activator protein or complex. TopBP1 activates the ATR-ATRIP complex.⁷ The MRN (Mre11-Rad50-Nbs1) complex stimulates the kinase activity of ATM, especially in the presence of double-stranded DNA ends.²⁰ The Ku70/Ku80 heterodimer when bound to DNA ends activates DNA-PKcs.²¹ Rheb stimulates the kinase activity of mTOR-Raptor complexes.²² The localization of each of the PIKKs to specific subcellular locations functions to concentrate the kinase complex with the protein activators.

Sequence variations in the PRD may allow PIKK family members to respond to different regulatory inputs. For ATR, the PRD is a binding surface for the ATR activator TopBP1.⁸ It will be important to determine whether the PRDs of other PIKKs are sites of interaction with other activator proteins. For example, we might expect the ATM PRD to interact with a subunit of the MRN complex. Rheb has been shown to bind to a C-terminal portion of mTOR that includes the PRD.²² Also, we have shown that the DNA-PKcs PRD is important for its activation.⁸ The DNA-PKcs PRD may be a point of contact between DNA-PKcs and the Ku70/80 complex. Besides serving as a binding domain for PIKK activators, the PRD is also a site for other regulatory inputs. Price and colleagues mapped a Tip60-catalyzed acetylation site to the ATM PRD.²³ Mutation of this acetylation site did not affect the basal kinase activity of ATM, but impaired the activation of ATM in response to double-strand breaks. It will be interesting to test if this acetylation mediates an interaction between ATM and the MRN complex or another ATM activator, such as Aven.²⁴ Finally, AKT targets mTORC1 directly at sites in the PRD, yet it is not clear how this regulates mTOR activity.²⁵

Kinetic Analysis of ATR Activation

How does the TopBP1 AAD regulate ATR kinase activity? It is unlikely that TopBP1 directly recruits substrates to the ATR kinase domain since TopBP1 increases the kinase activity of ATR towards all known substrates and even towards a non-physiological substrate PHAS-1.⁷ More likely would be a direct TopBP1 AAD effect on the conformation of the ATR kinase domain.

A TopBP1-dependent conformational change of the ATR kinase domain may increase its ability to bind ATP or substrates. To test this idea we varied the concentration of ATP and substrate in our ATR kinase assays. The stimulation of ATR kinase activity by TopBP1 is constant at different ATP concentrations (D.M., unpublished data). On the other hand, kinase activation varied considerably as a function of substrate concentration. ATR activation by TopBP1 was greatest at low substrate concentrations and declined exponentially as substrate concentration increased (Figure 1C). In other words, at saturating concentrations of substrate the effect of TopBP1 on ATR kinase activity is less pronounced. Estimations of the apparent K_m, which is inversely related to the affinity of an enzyme for its substrates, of ATR indicate that it is substantially reduced in the presence of TopBP1. Although kinase enzyme kinetics is complex due to the involvement of many biochemical steps, the key (rate-limiting) step that determines kinase specificity and activity is usually substrate binding. Therefore, simple steady state kinetics models may be used to gain insight into the substrate binding step. Our analysis suggests that TopBP1 affects the substrate binding step of the kinase reaction and increases the ability of ATR to binds its substrates. This may explain why we observe very modest changes in ATR-ATRIP autophosphorylation when TopBP1 is added to kinase reactions. Since ATRIP is already bound to ATR, from the perspective of the kinase domain it would appear as being at a relatively high concentration. Taken together, we speculate that the interaction between the ATR PRD and the TopBP1 AAD causes a conformational change in the kinase domain that relieves steric inhibition resulting in increased access of substrates to the substrate binding region of the catalytic site.

Figure 1. Functional Domains of ATRIP and ATR and Activation of ATR by TopBP1 as a function of substrate concentration.

(A) ATRIP contains an N-terminal Checkpoint Recruitment Domain (CRD), a coiled-coil domain that mediates it homo-dimerization, a TopBP1 interacting region, and a C-terminal ATR binding domain.⁴, ⁸, ²⁸, ²⁹ (B) ATR contains an N-terminal ATRIP binding domain.⁴ ATR contains a similar domain architecture to other PIKK family members. Its kinase domain is flanked by two regions of high sequence similarity named the FAT and FATC domains. The FAT domain is part of a long N-terminal region predicted to fold into an extended alpha-helical HEAT repeat structure.³⁰ The small FATC domain of at least some of the kinases is interchangeable.³¹ The FATC domain is required for ATR kinase activity.⁸ The functions of the FAT and FATC domain are unknown. Between the kinase domain and the FATC domain is the PRD (PIKK regulatory domain), which interacts with TopBP1, and is necessary for ATR activation.⁸ Sequence alignment of the PRD among ATR orthologs and ATM is presented. (C) In vitro ATR kinase assays were performed as described.⁸ The amount of the substrate, MCM2, was varied as indicated. Quantification of substrate phosphorylation was performed using a phosphorimager. Apparent K_m (μM) and V_max were calculated by nonlinear regression. Velocity is presented in arbitrary units.

Comparisons to PI3 Kinase

PI3Kα (phosphatidylinositol 3-kinase, isoform α catalytic subunit), which is frequently mutated in cancer, is composed of the p110α catalytic subunit and the p85 regulatory subunit. Recent structural determinations have shed light on its mechanism of regulation. The ultimate C-terminal α-helix (αK12) of the p110α kinase domain was shown to be in close proximity to both the catalytic activation loop and a domain of p85. This suggests that p85 contacts αK12 to induce a conformational change in the activation loop of p110α kinase domain, resulting in increased kinase activity.²⁶ Underscoring its importance in PI3Kα regulation, the αK12 is a hotspot for oncogenic mutations that result in a constitutively active kinase.²⁷ If the general mechanism of activation for PI3K lipid kinases is conserved with PIKK protein kinases, then the PRD may be functionally equivalent to αK12. The TopBP1 AAD may cause a PRD-mediated conformational change in the ATR kinase domain. Interestingly, the PRD of both ATR and ATM is predicted to contain a single α-helix. One prediction of this model would be that some mutations in the ATR PRD could result in ATR being locked in its active state. So far, we have not observed any PRD mutations that produce a constituvely active kinase. However, removal of a portion of the PRD in mTOR increases its basal kinase activity.²⁵

Conclusions and Future Directions

The PIK Kinases are key signal tranducers in controlling cellular responses to nutrients, RNA metabolism, and genotoxic stresses. The DNA damage responsive PIKKs may be useful drug targets for cancer since interfering with their activity causes chemo- and radio-sensitivity especially in tumor cells that already have defects in semi-redundant genome maintenance pathways. Our data suggest that there are common mechanisms of regulation of all the PIKKs. In particular, the PRD domain of the PIK kinases may translate specific regulatory inputs into a common conformational change that activates the kinases. Further delineating their activation mechanisms will require a more detailed kinetic analysis, identification of additional interacting partners, and the examination of post-translational modifications. Perhaps most importantly, a high-resolution structural determination of ATR or a related PIKK would provide tremendous insight into the precise mechanism of activation.

Abreviations

ATM

ataxia-telangiectasia mutated

ATR

ATM and Rad3-related

PIKK

phosphoinositide 3-kinase related kinase

PRD

PIKK regulatory domain

ATRIP

ATR-interacting protein

mTOR

mammalian target of rapamycin

DNA-PKcs

DNA-dependent protein kinase catalytic subunit

9-1-1

Rad9-Hus1-Rad1

MRN

Mre11-Rad50-Nbs1