nihms646777.pdf

Interaction of Chk1 with Treslin Negatively Regulates the

Initiation of Chromosomal DNA Replication

Cai Guo

Akiko Kumagai

Katharina Schlacher

2,4

Anna Shevchenko

Andrej

Shevchenko

, and

William G. Dunphy

1,*

Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA

91125, USA

Department of Molecular and Medical Pharmacology, University of California, Los Angeles, CA

90095, USA; Developmental Biology Program, Memorial Sloan-Kettering Cancer Center, New

York, NY 10065, USA

Max Planck Institute of Molecular Cell Biology and Genetics, 01307 Dresden, Germany

SUMMARY

Treslin helps to trigger the initiation of DNA replication by promoting integration of Cdc45 into

the replicative helicase. Treslin is a key positive-regulatory target of cell cycle control

mechanisms; activation of Treslin by cyclin-dependent kinase is essential for the initiation of

replication. Here we demonstrate that Treslin is also a critical locus for negative regulatory

mechanisms that suppress initiation. We found that the checkpoint-regulatory kinase Chk1

associates specifically with a C-terminal domain of Treslin (designated TRCT). Mutations in the

TRCT domain abolish binding of Chk1 to Treslin and thereby eliminate Chk1-catalyzed

phosphorylation of Treslin. Significantly, abolition of the Treslin-Chk1 interaction results in

elevated initiation of chromosomal DNA replication during an unperturbed cell cycle, which

reveals a function for Chk1 during a normal S-phase. This increase is due to enhanced loading of

Cdc45 onto potential replication origins. These studies provide important insights into how

vertebrate cells orchestrate proper initiation of replication.

INTRODUCTION

In eukaryotic cells, duplication of the genome depends upon the intricate, stepwise assembly

of protein complexes onto origins of DNA replication (

Sclafani and Holzen, 2007

;

Siddiqui

et al., 2013

;

Tanaka and Araki, 2013

). Initially, the origin recognition complex (ORC) and

Cdc6 associate with potential origins and thereupon recruit Cdt1 and the mini-chromosome

maintenance (MCM) complex. The MCM complex serves as the core of the replicative

Corresponding Author: dunphy@caltech.edu, Phone: (626) 395-8433, Fax: (626) 449-0756.

Present Address: Department of Cancer Biology, MD Anderson Cancer Center, Houston, TX 77054, USA

Publisher's Disclaimer:

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our

customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of

the resulting proof before it is published in its final citable form. Please note that during the production process errors may be

discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public Access

Author Manuscript

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

Published in final edited form as:

Mol Cell

. 2015 February 5; 57(3): 492–505. doi:10.1016/j.molcel.2014.12.003.

NIH-PA Author Manuscript

helicase that unwinds the DNA strands for replication. A key regulatory juncture in

replication involves the concerted binding of additional proteins to the MCM complex to

form the mature, activated version of the helicase. In particular, the Cdc45 and GINS

proteins associate with the MCM proteins and thereby form the CMG (Cdc45-MCM-GINS)

complex, which corresponds to the fully constituted helicase.

In vertebrates, integration of Cdc45 and GINS with the MCMs depends upon TopBP1 and a

recently discovered TopBP1-binding protein called Treslin (also known as Ticrr)(

Kumagai

et al., 2010

;

Sansam et al., 2010

). Importantly, formation of the TopBP1-Treslin complex

requires phosphorylation of Treslin by the S-phase cyclin-dependent kinase (S-CDK)(

Boos

et al., 2011

;

Kumagai et al., 2011

). Consequently, this phosphorylation helps to explain how

the cell cycle control system dictates the timing of S-phase. An analogous situation exists in

budding yeast where phosphorylation of the Treslin homologue Sld3 by S-CDK is also

critical for replication (

Labib, 2010

;

Siddiqui et al., 2013

;

Tanaka and Araki, 2013

Treslin is a relatively large protein (220 kD) that is approximately three times bigger than

yeast Sld3. Thus, Treslin may have acquired new properties that allow it to meet the more

complex demands of higher eukaryotes. To obtain further insight into established and

potentially novel functions of Treslin, we have engaged in a search for Treslin-interacting

proteins in human cells. These studies resulted in the identification of Chk1, an effector

kinase in checkpoint control mechanisms. Chk1 is best known for its role in blocking

activation of CDKs in cells with incompletely replicated or damaged DNA (

Perry and

Kornbluth, 2007

;

Toledo et al., 2011

Significantly, further studies have indicated that Chk1 also plays a role during a seemingly

normal cell cycle. For example, a number of observations have implicated Chk1 in the

control of replication during an unperturbed S-phase (

Maya-Mendoza et al., 2007

;

McIntosh

and Blow, 2012

;

Miao et al., 2003

;

Syljuasen et al., 2005

). The mechanism by which Chk1

exerts these effects is obscure. Chk1 also participates in suppressing endoreplication in

trophoblast stem (TS) cells (

Ullah et al., 2011

). Moreover, Chk1 appears to function in the

system that monitors correct attachment of chromosomes to the mitotic spindle (

Zachos et

al., 2007

). Overall, these observations suggest that Chk1 has diverse roles in cell cycle

regulation, which may help to explain why it is essential for viability (

Liu et al., 2000

Therefore, it will be important to understand how cells control the participation of Chk1 in

these varied functions.

In this report, we have investigated the molecular mechanism and functional consequences

of the Treslin-Chk1 interaction in both human cells and

Xenopus

egg extracts. We show that

Chk1 negatively regulates the Treslin-mediated loading of Cdc45 onto chromatin and

thereby serves to antagonize the initiation of replication. These studies provide an important

new perspective on how vertebrate cells control the initiation of DNA replication through

opposing negative and positive regulatory mechanisms. Furthermore, these experiments

reveal the mechanistic basis for a critical function of Chk1 apart from its role in checkpoint

responses to damaged DNA.

Guo et al.

Page 2

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

RESULTS

Chk1 Is a Treslin-interacting Protein

To search for Treslin-interacting proteins, we expressed various tagged versions of Treslin

in human 293T cells, re-isolated these polypeptides, and then analyzed associated proteins

by mass spectrometry (Figure 1A). For these experiments, we produced recombinant full-

length human Treslin with both S-peptide and 3X-FLAG tags at the C-terminal end

(designated Treslin-SF). We also prepared tagged fragments corresponding to residues

1-1257 and 1253-1909 of the protein. The 1-1257 fragment can restore DNA replication to

Treslin-depleted cells (

Kumagai et al., 2011

). Thus, the remaining C-terminal domain of

Treslin may have some regulatory role. For this study, we focused on proteins that might

bind selectively to this area.

We identified Chk1 as a protein that associated with full-length Treslin and the C-terminal

1253-1909 fragment, but not the N-terminal 1-1257 fragment (see Experimental Procedures

and Table S1). To validate these findings, we subjected S-protein pulldowns from cells to

immunoblotting with anti-Chk1 antibodies (Figure 1B). We likewise observed binding of

Chk1 to both the full-length protein and C-terminal fragment. Conversely, as expected from

previous studies, TopBP1 associated with the N-terminal but not C-terminal fragment

(

Kumagai et al., 2011

). We also performed reciprocal immunoprecipitation experiments in

human cells; we detected the presence of Treslin in anti-Chk1 immunoprecipitates and Chk1

in anti-Treslin immunoprecipitates, respectively (Figure 1C). Treatment with ethidium

bromide or Benzonase did not inhibit co-immunoprecipitation of Treslin and Chk1, which

rules out bridging of these proteins by DNA (Figure S1). Finally, to address whether this

binding required phosphorylation, we treated the anti-Chk1 immunoprecipitates with lambda

phosphatase. Although the phosphatase treatment appeared to be effective, as indicated by

increased electrophoretic mobility of Treslin, we observed no decrease in binding of Chk1 to

Treslin (Figure 1C). Overall, we conclude that Chk1 associates specifically with Treslin in

human cells.

Mapping of the Chk1-interacting Domain in Treslin

To investigate the molecular basis of this interaction, we set out to map a Chk1-interacting

region in Treslin. We engineered various sub-fragments of the 1253-1909 fragment,

expressed them in human cells, performed S-protein pulldowns, and immunoblotted for

Chk1. We found that approximately 100 amino acids at the C-terminal end of the protein

were necessary for binding to Chk1 (Figures 2A and 2B). We proceeded to show that a 100

amino acid fragment from the C-terminal end (residues 1810-1909) was also sufficient for

binding. The sequence of this region is well conserved in the

Xenopus

and zebrafish

homologues of Treslin (Figure 2C). Accordingly, we named this region the TRCT (

Treslin

C-

terminal) domain.

To characterize the TRCT, we prepared this domain as a GST fusion protein (Figure 2D).

The purified GST-TRCT bound well to Chk1 in human cell lysates. By contrast, GST-

tagged sub-fragments of this region (e.g., residues 1810-1872 and 1870-1909) did not

associate with Chk1 significantly, which suggests that the whole domain is necessary for

Guo et al.

Page 3

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

binding. The most conserved stretch within the human TRCT corresponds to the sequence

LTQSPLL at positions 1846-1852. This sequence is identical in

Xenopus

and zebrafish

homologues of Treslin, but is not present in budding yeast Sld3. We mutated each residue in

this sequence to alanine (to create the 7A mutant) in the context of both the GST-TRCT

construct and the full-length Treslin-SF protein. We found that 7A-mutant versions of these

polypeptides were completely defective for binding to Chk1 (Figures 2D and 2E). Notably,

the 7A mutant of full-length Treslin bound TopBP1 normally. We also identified three

residues within the TRCT (S1887, S1893, and T1987) that sit in consensus sequences for

phosphorylation by Chk1 (RXXS/T). However, mutation of any one of these residues to

alanine had no effect on the binding to Chk1 (Figure 2D). Finally, we assessed whether

Treslin associates with Chk1 directly. For this purpose, we incubated the TRCT domain with

purified, recombinant Chk1 (Figure S2). We observed that the WT TRCT bound to isolated

Chk1 very efficiently, whereas there was virtually no binding of the 7A mutant. Since no

phosphorylation could occur under these conditions, this observation reinforces the concept

that phosphorylation of the TRCT is not necessary for binding to Chk1. Overall, these

results indicate that Chk1 associates directly in a highly specific manner with sequences in

the C-terminal region of Treslin.

The TRCT Domain Promotes Chk1-catalyzed Phosphorylation of Treslin

One explanation for the binding of Chk1 to the TRCT would be that it facilitates

phosphorylation of Treslin by Chk1. To address this possibility, we first examined whether

Treslin could serve as a substrate of Chk1. We initially tested the GST-TRCT construct and

found wild-type (WT) recombinant Chk1 but not kinase-dead (KD) Chk1 could

phosphorylate this fragment well (Figure 3A). By contrast, there was no phosphorylation of

the mutant GST-TRCT-7A fragment by Chk1 above background levels. The S1893A and

T1897A mutants of the TRCT domain showed reduced phosphorylation by Chk1, while the

S1887A mutant was still an equally good substrate. A combined S1893A/T1897A mutant

displayed near background levels of phosphorylation by Chk1 (Figure S3A), which suggests

that these positions are the two main in vitro phosphorylation sites within this domain. As

anticipated, this mutant also still bound normally to Chk1 (Figure S3B). Phosphorylation of

the TRCT domain by Chk1 appears to be quite efficient. For comparison, phosphorylation of

a GST-tagged peptide from Cdc25 that contains a single well-documented site for Chk1

(

Kumagai et al., 1998

) was about 2-fold lower (Figure S3C). Finally, we likewise examined

full-length Treslin and found that the WT but not 7A-mutant protein could serve as good

substrate for Chk1 (Figures 3B and 3C). Taken together, these results indicate that docking

of Chk1 onto the TRCT domain strongly stimulates phosphorylation of Treslin by Chk1.

We employed mass spectrometry to identify in vitro phosphorylation sites for Chk1 on full-

length Treslin. This analysis resulted in the identification of numerous sites throughout

much of the protein (not shown). For these studies, we decided to focus on using the 7A

mutant to investigate the functional significance of the Treslin-Chk1 interaction. This

mutant is not an effective substrate for Chk1 in vitro and most likely in vivo. Furthermore,

this approach would also address the possibility that the physical association of Chk1 with

Treslin may also have a regulatory impact apart from phosphorylation.

Guo et al.

Page 4

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

Binding of Treslin Is Not Essential for Activation of Chk1

We previously demonstrated that ablation of Treslin from human cells compromises both

DNA replication and phosphorylation of Chk1 upon treatment with aphidicolin (APH)

(

Kumagai et al., 2010

). In principle, the latter defect could reflect a direct role for Treslin in

the activation of Chk1. Alternatively, this effect could be indirectly due to the absence of

replication forks in cells without Treslin. Because Treslin associates with Chk1, we asked

whether Treslin is directly necessary for activation of Chk1 in APH-treated cells. For this

purpose, we utilized lines of U2OS cells in which expression of siRNA-resistant versions of

WT and 7A Treslin was under the control of a doxycycline-inducible promoter (see

Experimental Procedures). To characterize these cell lines, we first used labeling with EdU

to assess whether the 7A mutant could rescue DNA replication in Treslin siRNA-treated

cells. We found that the percentages of EdU-positive nuclei were similar for Treslin-ablated

cells expressing WT or 7A Treslin (Figure 3D).

Next, we examined phosphorylation of Chk1 in cells treated with APH. As expected,

treatment of the parental U2OS cell line with APH in the presence of control siRNA

efficiently induced the phosphorylation of Chk1 on S345 (Figure 3E). As described

previously, treatment with Treslin siRNA resulted in markedly reduced phosphorylation of

Chk1. We proceeded to show that expression of either WT or 7A siRNA-resistant Treslin

could efficiently rescue phosphorylation of Chk1 in APH-treated cells that had been treated

with the Treslin siRNA. These observations indicate that the binding of Chk1 to the TRCT

domain is not essential for activation of Chk1 in response to APH.

The Isolated TRCT Domain of Treslin Stimulates DNA Replication in

Xenopus

Egg Extracts

We next considered the possibility that Chk1 regulates DNA replication by associating with

the TRCT domain. As one approach to examine this question, we utilized various types of

extracts from

Xenopus

eggs that recapitulate DNA replication in a cell-free reaction. In

particular, we utilized the nucleoplasmic extract (NPE) system, in which replication occurs

in a soluble nuclear fraction lacking membranes (

Walter and Newport, 2000

). We also used

whole egg extracts in which DNA replication takes place in reconstituted nuclei.

We reasoned that the isolated TRCT domain might act as a competitor of the interaction

between endogenous Treslin and Chk1 in egg extracts. To explore this possibility, we added

the GST-TRCT to the NPE system and then monitored the time-course of DNA replication.

For this purpose, we assessed incorporation of radioactive phosphate from [

P]dATP into

chromosomal DNA. We observed that addition of the TRCT fragment elicited a significant

increase in DNA replication in comparison with samples treated with control buffer alone or

the mutant TRCT-7A fragment (Figures 4A and 4B).

To assess whether this replication occurs by the normal CDK-mediated mechanism, we

utilized the CDK inhibitor p27 (Figure 4C). We noted that there was no DNA replication in

p27-treated extracts in either the presence or absence of the TRCT domain. To gauge

whether the TRCT does actually prevent the binding of Chk1 to endogenous Treslin, we

performed immunoprecipitation experiments with anti-Chk1 antibodies (Figure 4D). We

could readily detect Treslin in anti-Chk1 immunoprecipitates from NPE fractions.

Guo et al.

Page 5

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

Association of Treslin with Chk1 was not inhibited by treatment with ethidium bromide or

Benzonase, agents that would prohibit bridging by DNA (Figure S4A). We observed that

addition of the WT TRCT fragment to NPE fractions caused a severe reduction in the

binding of Treslin to Chk1, whereas the 7A mutant fragment had no effect. By contrast, the

TRCT domain had no effect on the binding of Treslin to TopBP1. Overall, these results

indicated that blockage of the binding of Chk1 to Treslin results in elevated DNA

replication.

Chk1 Regulates DNA Replication in

Xenopus

Egg Extracts

As another means to investigate the regulation of DNA replication by Chk1, we utilized the

Chk1 inhibitor AZD7762. We found that this compound also elicited an increase in DNA

replication comparable to that induced by the TRCT-WT fragment (Figure 4E and Figure

S4B). We also attempted to immunodeplete Chk1 from the NPE system, which entails

separate immunodepletion from the HSS and NPE fractions that are necessary for these

experiments. However, these immunodepletion procedures resulted in a non-specific

reduction of replication, which confounded this approach.

As an alternative, we attempted to deplete Chk1 from whole egg extracts (

Kumagai et al.,

1998

). We first needed to assess whether the TRCT affects replication in whole egg extracts,

which would require transport of the TRCT into reconstituted nuclei in these extracts. Since

our initial GST construct lacked a nuclear localization sequence (NLS), we prepared a new

construct with this sequence (GST-TRCT-NLS) (Figure 5A). We verified that both WT and

7A versions of the TRCT-NLS could be found in lysates of reconstituted nuclei from whole

egg extracts. Unexpectedly, the WT but not 7A version of the TRCT-NLS also caused a

dramatic accumulation of Chk1 in the nuclei (Figure 5B). A plausible explanation is that the

TRCT-NLS, because of its robust binding to Chk1, may promote nuclear entry of Chk1.

Next, we added the WT TRCT-NLS to whole egg extracts containing reconstituted nuclei

and found that this fragment also elicited an increase in DNA replication relative to

incubations containing control buffer or a 7A mutant version of the TRCT-NLS (Figures 5C

and 5D). Finally, we removed Chk1 from whole egg extracts with anti-Chk1 antibodies

(Figure 5E). We observed a significant acceleration of DNA replication in Chk1-depleted

extracts (Figures 5F and 5G). Furthermore, addition of recombinant Chk1 to the depleted

extracts restored replication to the lower level found in mock-depleted extracts. Taken

together, these experiments provide multiple lines of evidence that Chk1 regulates DNA

replication in the

Xenopus

system.

The Isolated TRCT Domain Promotes Initiation of Replication in Egg Extracts

A key function of Treslin involves the loading of Cdc45 onto chromatin (

Kumagai et al.,

2010

). It has been well established that the loading of Cdc45 is critical for the initiation of

DNA replication in both yeast and vertebrates (

Sclafani and Holzen, 2007

;

Siddiqui et al.,

2013

;

Tanaka and Araki, 2013

). To explore the basis of the TRCT-mediated stimulation of

replication, we examined the loading of Cdc45 onto chromatin in the NPE system in the

presence of this fragment. We observed that addition of the TRCT-WT protein stimulated a

large increase in the loading of Cdc45 onto chromatin in comparison with samples

Guo et al.

Page 6

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

containing the TRCT-7A protein (Figure 6A). There was a similar increase in the loading of

Sld5 (a component of GINS), PCNA, and DNA polymerase epsilon (Figure S5). The TRCT

domain typically did not affect the binding of the ORC and MCM complexes to the DNA.

These experiments suggest that the isolated TRCT domain promotes initiation by

stimulating the loading of the helicase activators Cdc45 and GINS onto chromatin.

Furthermore, we also observed an increase other replication-fork proteins, such as PCNA

and DNA polymerase epsilon, which participate directly in the ensuing DNA synthesis.

As another method to characterize this phenomenon, we attempted to block replication just

after initiation. Actinomycin D blocks replication in the NPE system shortly after initiation

but before significant unwinding of the DNA (

Pacek and Walter, 2004

). Accordingly, we

added actinomycin D to the NPE system in the absence and presence of the TRCT domain.

We found that actinomycin D caused a pronounced further accumulation of Cdc45 on the

DNA in extracts containing the WT TRCT domain (Figure 6B). There was no effect in

extracts containing the TRCT-7A mutant. Taken together, these experiments suggest that the

TRCT domain promotes the initiation of replication. The fact that the TRCT domain blocks

the binding of Chk1 to Treslin suggests that Chk1 suppresses the initiating function of

Treslin.

To obtain further support for this concept, we performed DNA fiber studies in whole egg

extracts following addition of the TRCT domain. For this purpose, we incubated extracts

sequentially with digoxigenin-dUTP and biotin-16-dUTP, prepared DNA fibers from

nuclear fractions, and then measured inter-origin distances (

Bellelli et al., 2014

;

Marheineke

et al., 2009

). We observed that the WT TRCT fragment caused a significant increase in

shorter inter-origin distances (e.g., 10–15 kb) relative to incubations containing added buffer

alone or the 7A fragment (Figure 6C; Figures S6A and S6B). Concomitantly, there was a

decrease in larger inter-origin distances (e.g., >30 kb) in incubations containing the WT

fragment. These findings directly support an increase in the firing of origins, which fits well

with the observations that the TRCT domain enhances formation of the activated replicative

helicase.

Dysregulated Origin Firing in the Absence of the Treslin-Chk1 Interaction Induces

Activation of Chk1

The elevated loading of Cdc45 onto chromatin and the ensuing increase in replication in the

presence of the TRCT domain raised the possibility that this fragment might elicit

replication stress. To address this issue, we examined phosphorylation of

Xenopus

Chk1 on

S344 (

Kumagai et al., 1998

). We observed that the WT TRCT domain induced

phosphorylation of Chk1 in NPE fractions containing sperm chromatin even in the absence

of APH (Figure 6D). The TRCT domain also caused a substantial further increase in

phosphorylation of Chk1 in extracts containing APH. By contrast, there was no increase in

extracts treated with control buffer or the 7A mutant TRCT domain. Furthermore, the

increase in the presence of the WT TRCT domain was abolished by caffeine, an inhibitor of

the ATR-catalyzed phosphorylation of Chk1. Therefore, the TRCT-stimulated activation of

Chk1 involves the ATR-mediated pathway. This increase also depended on the presence of

sperm chromatin as a template for replication, which suggests that the TRCT domain does

Guo et al.

Page 7

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

not somehow directly activate Chk1 by a DNA-independent mechanism. In further support

of this concept, we found that inhibition of DNA replication with p27 abolished the TRCT-

stimulated increase in the phosphorylation of Chk1 in extracts lacking APH (Figure S6C).

Overall, these experiments indicate that the isolated TRCT domain causes a pronounced

derangement of DNA replication in the

Xenopus

egg extract system.

Expression of the Treslin-7A Mutant in Human Cells Leads to Increased Origin Firing

As another strategy to examine the function of the Treslin-Chk1 interaction, we utilized

human cells that overexpress the Treslin-7A mutant. In particular, we employed U2OS T-

REx cells that express WT and 7A Treslin in a doxycycline-inducible manner. We induced

the expression of Treslin in these cells by addition of doxycycline and later examined origin

firing by DNA fiber analysis (

Jackson and Pombo, 1998

;

Schlacher et al., 2011

). For these

studies, we incubated the cells with CldU for 20 min and then with IdU for 20 min (Figure

7A). Next, we prepared DNA fibers from the cells by standard methods and detected

incorporation of the modified nucleotides with fluorescently tagged antibodies that detect

CldU (green tracks) and IdU (red tracks). Finally, we quantitated the frequency of singly

labeled IdU (red) tracks as a measure of new origin firing during the second labeling period.

We noted that there was a significant increase in origin firing in cells expressing the full-

length Treslin-7A mutant in comparison with cells expressing the WT protein (Figures 7B–

7D). The induced levels of the WT and 7A proteins were quite similar in the different cell

lines. Moreover, overexpression of WT Treslin did not have a significant effect on origin

firing. We also examined replication fork speed in these cells but could not discern a

difference between cells expressing WT or 7A Treslin (Figure S7). Overall, these

experiments further support our findings with

Xenopus

egg extracts that disruption of the

Treslin-Chk1 interaction leads to increased origin firing. The fact that the Treslin-Chk1

interaction is functionally important in both

Xenopus

and humans suggests that this

regulatory mechanism is a conserved feature of DNA replication in vertebrates.

DISCUSSION

In this report, we have identified an interaction between Treslin and the key checkpoint-

effector kinase Chk1. The high specificity of this binding suggested that there is a significant

regulatory relationship between these two proteins. Indeed, we have found that Chk1

negatively regulates the replication-initiating function of Treslin (see Figure 7E). To reach

this conclusion, we utilized both

Xenopus

egg extracts and human cells. In one approach, we

added the TRCT domain from Treslin to

Xenopus

egg extracts as a competitor of the

Treslin-Chk1 interaction. Strikingly, this peptide elicited a pronounced increase in DNA

replication.

In searching for the basis of this phenomenon, we found that the TRCT fragment strongly

stimulated the binding of Cdc45 to chromatin. The loading of Cdc45 onto chromatin in egg

extracts is rate limiting for origin firing (

Mimura et al., 2000

;

Walter and Newport, 2000

Thus, the TRCT domain appears to act by enhancing initiation. To provide further evidence,

we have dissected this process with actinomycin D, a drug that blocks replication in egg

extracts just after initiation but before significant unwinding of the DNA (

Pacek and Walter,

Guo et al.

Page 8

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

2004

). Treatment with both the TRCT domain and actinomycin D led to a dramatic and

sustained accumulation of Cdc45 on chromatin. Finally, we utilized DNA fiber studies to

show that the addition of the TRCT domain to egg extracts leads to an overall decrease in

inter-origin distances in replicating chromatin.

In a complementary approach, we also employed human cells. In particular, we also used

DNA fiber analysis to show that human cells overexpressing the full-length Treslin-7A

mutant display increased origin firing during an unperturbed S-phase. This observation is

consistent with the fact that human cells with compromised function of Chk1 display

elevated firing of replication origins during a normal cell cycle (

Maya-Mendoza et al., 2007

;

McIntosh and Blow, 2012

;

Miao et al., 2003

;

Syljuasen et al., 2005

). Hence, our studies

have revealed a mechanism by which Chk1 controls replication in cells without overtly

damaged DNA. It has also been reported that Chk1-deficient cells display decreased

replication-fork speed (see

Petermann et al., 2010

). We have not observed an obvious effect

on fork speed in cells expressing the Treslin-7A mutant. This mutant may not cause enough

origin firing to disrupt fork progression. Moreover, Chk1 could have distinct targets for

initiation versus elongation.

Utilization of replication origins places a number of regulatory demands on cells. The firing

of origins depends upon positive regulation of Treslin by S-CDK, but there may also have to

be negative regulatory mechanisms that suppress premature or inappropriate action of

Treslin at origins. The ability of Chk1 to inhibit the Treslin-mediated loading of Cdc45

would provide such a mechanism. This process could operate generally at all origins or

come into play at a subset of origins under certain circumstances.

Typically, the number of loaded MCM complexes on chromatin at the onset of S-phase

greatly exceeds the number of origins that actually fire in a given cell cycle. This “MCM

paradox” has suggested that there are a large number of dormant origins in the genome (see

McIntosh and Blow, 2012

). It has been proposed that utilization of these dormant origins

would allow cells to cope with replication stress by increasing the likelihood that replication

in stressed areas would reach completion. The question arises, however, about how the cell

would regulate the firing of such dormant origins under both unstressed and stressed

conditions. It has been postulated that these origins might fire stochastically such that

passive replication from nearby origins would typically occur under unstressed conditions.

However, under stressed conditions of slowed or blocked replication, there would be more

time for the dormant origins to undergo firing.

Another type of explanation would incorporate the existence of an inhibitory mechanism

that suppresses firing of dormant origins. In principle, inhibition of Treslin-mediated loading

of Cdc45 by basally active Chk1 under unstressed conditions could correspond to such a

mechanism. Our DNA fiber studies with egg extracts do not suggest that there is rampant

firing of dormant origins upon addition of the TRCT domain. However, more limited firing

of dormant origins may contribute to the accelerated replication that we have observed under

these conditions.

Guo et al.

Page 9

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

The Treslin-Chk1 interaction might also help to explain the temporal order in firing of

origins during S-phase in somatic cells. In eukaryotic cells, there is typically a pattern

wherein “early” and “late” origins at distinct chromosomal regions fire at different times

(

McIntosh and Blow, 2012

;

Siddiqui et al., 2013

). In budding yeast, regulation of the Treslin

homologue Sld3 has a role in the distinction between early and late origins (

Lopez-

Mosqueda et al., 2010

;

Zegerman and Diffley, 2010

). However, the mechanisms that

regulate such differential firing in vertebrates are largely unknown. Another potential

explanation for the inhibition of Treslin by Chk1 is that this process could be necessary for

the suppression of later-firing origins during the earlier parts of S-phase. Further studies will

be required to identify the characteristics of origins that are subject to Chk1-dependent

inhibition of Treslin.

It is well accepted that the ATR-dependent activation of Chk1 in response to incompletely

replicated and damaged DNA leads to inhibition of origin firing at chromosomal regions that

are not already engaged in replication (

McIntosh and Blow, 2012

;

Sorensen and Syljuasen,

2012

;

Yekezare et al., 2013

). The activation of Chk1 results in decreased function of Cdc25

and increased function of Wee1 (

Perry and Kornbluth, 2007

). In turn, these effects promote

the inhibitory phosphorylation of Cdk2 (as well as Cdk1) on T14 and Y15. The loading of

Cdc45 onto chromatin requires phosphorylation of Treslin by Cdk2, which mediates binding

of Treslin to TopBP1 (

Boos et al., 2011

;

Kumagai et al., 2011

). Therefore, it might be

anticipated that reduced activity of Cdk2 would compromise formation of the Treslin-

TopBP1 complex. Indeed, treatment of human cells with hydroxyurea reduces the binding of

Treslin to TopBP1, and the Chk1 inhibitor AZD7762 reverses this effect (

Boos et al., 2011

We have not been able to detect an override of the hydroxyurea-induced block to replication

in U2OS cells expressing the Treslin-7A mutant (not shown). This mutation would

presumably not affect the ability of Chk1 to down-regulate Cdk2. We have also observed

that both the WT and 7A forms of Treslin bind equally well to TopBP1. Moreover, the

Treslin-7A mutant displays normally regulated binding to TopBP1 (i.e., decreased binding

in the presence of hydroxyurea) (not shown). Finally, the TRCT domain does not affect

binding of Treslin to TopBP1 in

Xenopus

NPE fractions. It would be interesting to examine

the effect of the Treslin-7A mutant in replication-stressed cells in which the inhibition of

Cdk2 had been overridden.

Although inhibition of Cdk2 could account for the effect of Chk1 on DNA replication in

stressed cells, it is unclear that this mechanism could explain the role of Chk1 in unstressed

cells. The activity of Cdk2 rises substantially at S-phase in unperturbed cells. It is possible

that there could be some local inhibition of Cdk2 in certain regions of the genome, but the

mechanistic basis for such an effect is unknown. Another explanation is that Chk1 could

control replication in some other manner besides regulation of Cdk2 under conditions where

there is not a strong, exogenous threat to the DNA. The regulation of Treslin by Chk1 may

be such a mechanism. It should also be noted that our studies do not rule out the possibility

that Chk1 could have an additional target(s) besides Treslin during an unperturbed S-phase.

Overall, inhibition of Treslin by Chk1 could suppress relevant origins during an unperturbed

S-phase and perhaps even supplement the inhibition of Cdk2 in replication-stressed cells.

Guo et al.

Page 10

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript

A previous study in the

Xenopus

system indicated that depletion of Chk1 did not increase

DNA replication in either untreated or APH-treated egg extracts (

Luciani et al., 2004

). For

the untreated extracts, these investigators quantitated DNA replication at a single late time

point (120 min) when replication would have reached completion. Hence, this analysis

would not have detected the early acceleration of replication that we have observed in this

study. Another group found that the Chk1 inhibitor UCN-01 did not increase replication in

egg extracts, but actually inhibited replication to some extent (

Murphy and Michael, 2013

However, UCN-01 also inhibits Cdk2 at the concentrations used in this study (

Kawakami et

al., 1996

), which complicates interpretation of the results.

In principle, phosphorylation of Treslin by Chk1 may alter its conformation or directly

affect its interactions with other proteins to preclude helicase activation. Moreover, the

physical association of Chk1 might also have such effects on Treslin. Regulation by Chk1

appears not to reduce binding of Treslin to TopBP1, but Chk1 may nonetheless alter the

helicase-activating properties of this complex. Chk1 may also inhibit the ability of Treslin

and TopBP1 to recognize Cdc45/GINS effectively or deliver Cdc45/GINS to the MCM

complex. Finally, Chk1 may influence the interaction of Treslin with some other

component(s) of the replication apparatus. We will need to understand better the exact

molecular mechanism by which Treslin promotes helicase activation in order to evaluate

which, if any, of these possibilities is correct.

In conclusion, we have used both the

Xenopus

and human systems to provide new

perspectives on the function of Chk1 and the regulation of early steps in DNA replication.

Further analysis of this process may yield additional insights into how vertebrate cells

ensure the faithful propagation of their genomes.

EXPERIMENTAL PROCEDURES

Plasmids

The cDNA for human Treslin (GenBank ADC30133.1) was described before (

Kumagai et

al., 2010

). pcDNA5/TO-Treslin-SF (encoding the SV40 NLS, an S-peptide tag, and a 3X-

FLAG tag at the C-terminal end of Treslin) was created from pcDNA5/TO-Treslin-Myc by

using PCR to exchange tags (

Kumagai et al., 2011

Human Tissue Culture Cells

293T and U2OS cells were cultured in DMEM containing 10% fetal bovine serum. 293T

cells were transfected using FuGENE6. Large-scale transfections as well as preparation of

nuclear lysates are described in Supplemental Experimental Procedures.

Antibodies

Anti-Treslin antibodies were previously described (

Kumagai et al., 2010

). For other

antibodies, see Supplemental Experimental Procedures.

Guo et al.

Page 11

Mol Cell

. Author manuscript; available in PMC 2016 February 05.

NIH-PA Author Manuscript