Báo cáo Y học: Distinct parts of minichromosome maintenance protein 2 associate with histone H3/H4 and RNA polymerase II holoenzyme pptx

Distinct parts of minichromosome maintenance protein 2 associate with histone H3/H4 and RNA polymerase II holoenzyme Linda Holland, Michael Downey, Xiaomin Song*, Laura Gauthier, Patrici

Distinct parts of minichromosome maintenance protein 2 associate with histone H3/H4 and RNA polymerase II holoenzyme

Linda Holland, Michael Downey, Xiaomin Song*, Laura Gauthier, Patricia Bell-Rogers

and Krassimir Yankulov

Department of Molecular Biology and Genetics, University of Guelph, Ontario Canada

Minichromosome maintenance (MCM) proteins are part of

the replication licensing factor (RLF-M), which limits the

initiation of DNA replication to once per cell cycle We have

previously reported that higher order complexes of

mam-malian pol II and general pol II transcription factors,

referred to as pol II holoenzyme, also contain MCM

pro-teins In the present study we have analyzed in detail the

interaction between MCM2 and pol II holoenzyme N- and

C- terminal deletions were introduced into epitope-tagged

MCM2 and the truncated proteins were transiently

expressed in 293 cells Affinity chromatography was used to

purify RNA pol II holoenzyme and histone binding MCM

complexes We found that amino acids 168–230 of MCM2 are required for its binding to pol II holoenzyme in vivo We also showed that bacterially expressed amino acids 169–212

of MCM2 associate with pol II and several general tran-scription factors in vitro Point mutations within the 169–212 domain of MCM2 disrupted its interaction with pol II holoenzyme both in vitro and in vivo This region is distinct from the previously characterized histone H3 binding domain of MCM2

Keywords: MCM2; RNA polymerase II holoenzyme; his-tone

Large protein complexes, which contain RNA polymerase

II as well as the general pol II transcription factors (GTFs)

TFII A, B, D, E, F, and H [1] and other proteins have been

isolated from yeast, mammalian and amphibian cells [2–8]

They are referred to as pol II
holoenzyme Some of the

components of pol II holoenzyme make direct or indirect

contacts with the C-terminal domain (CTD) of the largest

subunit of pol II Antibodies against the CTD disrupt the

yeast holoenzyme into core pol II and a mediator

subcom-plex, which contains the SRB and MED proteins [7–9] A

similar treatment of pol II holoenzyme from HeLa cells also

disrupts its interaction with several of the GTFs [6] In

higher eukaryotes the CTD mediates the interaction with

complexes that contain homologues of the yeast SRB and

MED proteins such as SMCC [10] or NAT [11] It is

believed that pol II holoenzyme is a functionally significant

complex, which is responsible for transactivator-stimulated

transcription in vivo It has been shown that the srb4 and

srb6genes are essential for expression of most mRNAs in budding yeast [12] Other holoenzyme components such as SRB 2, 5, 7–11, SWI/SNF proteins, SIN4, RGR1, MED2, MED9/CSE2, MED10/NUT2, MED11, GAL11, PGD1 and ROX3 [7,9,13–16] are not essential for transcription of most genes but do contribute to the response to transacti-vators and repressors (reviewed in [17,18]) In addition to its role in the response to transcriptional regulators, pol II holoenzyme may be involved in integrating transcription with RNA processing, DNA repair and replication In support of this idea, the DNA repair factors DNA pol e, XPC, XPF, XPG, Ku, RAD51 [3], BRCA1 [19]; RNA helicase A [20]; the replication factors RP-A, RP-C [3] and MCM proteins [6]; and the cleavage/polyadenylation factors CPSF and CstF [21] have been identified in mammalian pol II holoenzyme preparations There are significant differences in the composition of pol II holoen-zymes that have been purified by different procedures indicating that this complex is capable of interacting with a variety of proteins and that there might be multiple forms of pol II holoenzyme in vivo [22]

MCM proteins, previously characterized as components

of the replication licensing factor M (RLF-M) [23–26], were also found to associate with pol II holoenzyme in higher eukaryotes [6] It is believed that RLF-M is acting to limit replication of genomic DNA to a single round per cell cycle [27] As predicted by the licensing model, most MCMs are released from chromatin during S phase and re-associate at the end of mitosis [28–33] In addition to promoting initiation of DNA replication, MCMs also seem to function

in replication fork movement [28,34,35] The MCM4,6,7 subcomplex possesses DNA helicase activity [35–38], which has been implicated in both initiation and fork movement

In addition, MCM complexes bind with high affinity to core histone H3/H4 dimers [39,40] and to HBO1 [41], via distinct domains in the N-terminus of MCM2 indicating a possible

Correspondence to K Yankulov, Department of Molecular Biology

and Genetics, University of Guelph, Guelph, Ontario N1G 2W1,

Canada Fax: + 1 519 8372075, Tel.: + 1 519 8244120, ext 6466,

E-mail: yankulov@uoguelph.ca

Abbreviations: MCM, minichromosome maintenance; CTD,

carboxy-terminal domain (of the largest subunit of RNA

poly-merase II); GTF, general transcription factor; SMCC, SRB/

MED-containing cofactor; NAT, negative regulator of activated

transcription; RLF-M, replication licensing factor M; TBP, TATA

box binding protein; HBO1, histone acetyltransferase binding to

ORC; ORC, origin recognition complex; FCS, fetal calf serum;

GST, glutathione S-transferase; TFIIS, transcription factor II S.

*Present address: Pharmacia Corporation, AA215/AA2C,

700 Chesterfield Parkway, Chesterfield, MO 63198, USA.

(Received 11 June 2002, revised 26 August 2002,

accepted 30 August 2002)

chromatin remodeling function MCM2, but not other

MCM proteins, also interacts with cdc6, a component of the

replication preinitiation complex [42] The significance of

these protein interactions and the precise biochemical role of

MCMs in regulating DNA replication remain unclear

There are certain indications that MCMs might be

involved in pol II transcription in higher eukaryotes We

have shown that antibodies against MCM2 inhibit pol II,

but not pol III, transcription from injected template

plasmids in Xenopus oocytes [6] Two other studies have

demonstrated that an interaction between MCM5 and

the activation domain of Stat1a is essential for the

expression of IFN-c responsive genes [43,44] On the

other hand, recruitment of pol II holoenzyme to origins

of replication via GAL11 or TBP significantly stimulates

replication of minichromosomes in S cerevisiae [45],

suggesting a possible role of pol II holoenzyme in

DNA replication

In this paper we have analyzed the interaction between

human MCM2 (also called BM28) and pol II holoenzyme

We report that MCM2 binds to pol II holoenzyme via a

sequence in its N-terminal domain This region is positioned

between the site of interaction with histone H3 and the

putative HBO1 binding site

M A T E R I A L S A N D M E T H O D S

Plasmids

All plasmids for expression of recombinant human

MCM2 encode N-terminally FLAG-tagged polypeptides

pFLAG-MCM2(FL) contains the EcoRI fragment of

pBSBM28 ([46], EMBL accession no P49736), cloned

into the EcoRI site of pFLAG-CMV-2 (Sigma) This

plasmid and all its derivatives encode MDYKDDDDK

LAAANSAESSESFT followed by different MCM2

frag-ments pFLAG-MCM2(1–197), pFLAG-MCM2(1–247),

and pFLAG-MCM2(1–511) were generated by deleting

the Sal I, the Bgl II, and the EcoRV fragments from

pFLAG-MCM2(FL), respectively pFLAG-MCM2(1–

167) was generated by subcloning the EcoR1-Dra III

fragment of pBSBM28 into EcoRI-Sma I linearized

pFLAG-CMV-2 pFLAG-MCM2(1–230) was generated

by deleting the BsaAI-Sma I fragment from

pFLAG-MCM2(1–247) pFLAG-MCM2(198–892) contains the

SalI fragment of pBSBM28 [46] cloned into the Sal I

site of pFLAG-CMV-2 and encodes MDYKDDDDK

LAAANSSIDLISVPV followed by amino acids 198–892

pFLAG-MCM2(345–892) contains the FseI-Sma I

frag-ment of pBSBM28 [46] cloned into NotI-Sma I linearized

pFLAG-CMV-2 and encodes MDYKDDDDKLA

fol-lowed by amino acids 345–892 of MCM2 All expression

plasmids were purified by anion exchange (Qiagen) prior

to transfection pGEX-MCM2(169–212) contains the

sequence encoding amino acids 169–212 of MCM2

attached in frame to GST Site-directed mutagenesis of

the pGEX-MCM2(169–212) and pFLAG-MCM2(1–230)

plasmids was conducted using the Quikchange

site-direc-ted mutagenesis kit (Stratagene) 5¢-CCGCTTCAA

GAACTTCCCGGGCACTCACGTCAC-3¢ was used as

a primer to introduce changes from LR to PG at

positions 192/193, and 5¢-GCCACGGCCACAACGAG

CTCAAGGAGCGCATCAGC-3¢ was used to introduce

changes from VF to EL at positions 203/204 Point mutations were confirmed by nucleotide sequencing Antibodies

Anti-(Pol II CTD) (8WG16) [47], anti-(BM28-N) directed towards the N-terminus of MCM2 [32], TBP [6], anti-TFIIB [6], anti-CPSFp160 [6], and anti-(Xenopus ORC2) [48] were described previously Anti-MCM was generated against a highly conserved peptide sequence VVCI DEFDKMSDMDRTA, which is shared between all MCM proteins [49] The antibody was affinity purified on antigenic peptide coupled to Affigel-10 (Bio-Rad) Anti-p62(TFIIH) was raised against full length human p62 expressed in E coli The anti-FLAG antibody M2 was purchased from Sigma The anti-CycC was from Santa Cruz Biotechnology

Expression of recombinant MCM2 proteins in human embryonic kidney fibroblast cells (293)

HEK 293 cells were grown in 15 cm plates (Costar) to 50% confluency in DMEM medium supplemented with 10% fetal bovine serum and antibiotics (100 unitsÆmL)1penicillin and 100 lgÆmL)1streptomycin) Each plate was transfected with 20 lg of pFLAG-CMV-MCM2 expression plasmid plus 20 lg of carrier plasmid (pBS) using calcium phos-phate-precipitation Transfection efficiency was between 30% and 50% as monitored by the expression of green fluorescent protein from pEGFP-C2 (Clontech)

Preparation of whole-cell extract Cells were harvested 36–48 h after transfection and whole-cell extract was prepared by lysis in hypotonic buffer and 0.41 M (NH4)2SO4 extraction as described previously [6] Prior to chromatography, each extract was buffer exchanged in a 10DG column (Bio-Rad) to chromato-graphy buffer (CB) (10 mM Hepes 7.9, 0.2 mM EDTA, 0.2 mMEGTA, 5 mM2-glycerophosphate, 1 mMNa3VO4,

1 mMNaF, 1 mMbenzamidine, 1 mMdithiothreitol, 50 lM ZnCl2, 1 lg of pepstatinÆmL)1, 1 lg of leupeptinÆmL)1,

2 lg of aprotininÆmL)1, 12% glycerol, 0.05% NP-40) plus

50 mMNaCl and clarified by centrifugation for 15 min at

21 000 g Protein concentration of the extracts after dialysis was 5–10 mgÆmL)1

GST-TFIIS affinity chromatography RNA polymerase II holoenzyme was purified by GST-TFIIS affinity chromatography as described previously [6] Briefly, Glutathione S-transferase (GST), and GST-TFIIS (residues 1–301 of mouse transcription factor IIS) were expressed in E coli BL21(LysS) and immobilized on Glutathione Sepharose 4B (Pharmacia) at 10 mgÆmL)1 Mini-columns containing 100 lL of beads were prepared Extracts from four tissue culture plates were passed through

a GST column followed by a GST-TFIIS column Each column was washed twice with 1 mL of CB plus 50 mM NaCl, then five times with 100 lL of CB plus 50 mMNaCl, eluted four times with 100 lL CB plus 0.325MNaCl and then four times with 100 lL with CB plus 1MNaCl Final

100 lL wash and eluate fractions were precipitated in

0.8 mgÆmL)1deoxycholic acid and 20% trichloroacetic acid

and then re-suspended in SDS/PAGE sample loading

buffer

Histone H3/H4 affinity chromatography

H3/H4 histones were purified from HeLa cell nuclear pellets

following the protocol of Simon and Felsenfeld [50] and

coupled to Affigel-10 (Bio-Rad) at 5 mgÆmL)1 of resin

Bovine serum albumin (Fraction V, Sigma) was coupled to

Affigel-10 at a concentration of 5 mgÆmL)1 Purification of

MCM proteins on H3/H4-Affigel beads was carried out as

described previously [39] with some modifications Briefly,

flow-through fractions from the GST-TFIIS columns

(3 mL at 5 mgÆmL)1protein) were loaded sequentially to

a BSA-Affigel column (100 lL) followed by a histone

H3/H4 column (100 lL) equilibrated with buffer A (20 mM

Tris/HCl pH 7.5, 0.5 mM EDTA, 1 mM dithiothreitol,

0.1 mMphenylmethanesulfonyl fluoride, and 10% glycerol)

containing 0.1M NaCl The columns were washed two

times with 1 mL and fiv e times with 200 lL of buffer A plus

0.1MNaCl, and were eluted with 0.5M, 0.75M and 2M

NaCl in buffer A (1 mL, 600 lL, and 2 mL, respectively)

Wash and eluate fractions were precipitated in 0.8 mgÆmL)1

deoxycholic acid/20% trichloroacetic acid, then

resuspen-ded in SDS/PAGE sample loading buffer

GST-MCM2(169–212) affinity chromatography

GST-MCM2(169–212)L192P/R193G, and GST-MCM2(169–

BL21(LysS)DE3 cells and coupled to glutathione Sepharose

4B (Pharmacia) at 10 mgÆmL)1 Each column (250 lL) was

loaded with HeLa whole cell extract (10 mgÆmL)1 [6]),

columns were extensively washed and eluted with 1MNaCl

Samples were precipitated with 0.8 mgÆmL)1deoxycholic

acid and 20% trichloroacetic acid and analyzed by Western

blotting

Western blotting

Proteins were transferred to Immobilon-P membrane

(Mil-lipore) by semidry electroblotting Blots were developed by

BM Chemiluminescence Blotting Substrate (Roche) or ECL

Plus (Amersham) with horseradish peroxidase coupled to

secondary antibody (Sigma or Amersham) For

quantita-tion, blots were exposed on an Image Station (Kodak,

440CF) and images were analyzed by Kodak 1D Image

Analysis Software

Proteomics tools

Multiple sequence analysis was performed by BLAST

Three dimensional structure prediction was carried out by

3D-PSST (http://www.bmm.icnet.uk/) and Swiss-Model

(http://www.expasy.ch/swissmod/SWISS-MODEL.html)

Prediction of sites of phosphorylation was byNETPHOS2.0

(http://www.cbs.dtu.dk/services/NetPhos) Hydrophobicity

and charge analysis was performed by PROTPARAM

(http://www.expasy.ch/tools/protparam.html) Secondary

structure prediction was by JPRED2 (http://jura.ebi.ac

uk:8888)

R E S U L T S

Experimental Strategy Previous experiments have shown that antibodies against MCM2 specifically inhibit pol II transcription in Xenopus oocytes [6] We decided to search for domain(s) in this polypeptide that might be responsible for the interaction between MCM proteins and pol II holoenzyme [6] FLAG-tagged human MCM2 deletion mutants (Fig 1) were expressed in 293 cells and assayed for their ability to copurify with pol II holoenzyme or to bind to histones H3/H4 This approach circumvented problems with bac-terial expression of MCM2 that we had encountered in the past (data not shown) Extracts were prepared from transfected cells and pol II holoenzyme was purified by affinity chromatography using GST-TFIIS as a ligand [5,6]

We had previously shown that about 2% of the total endogenous MCM2 in HeLa cell extract copurified with pol II holoenzyme on GST-TFIIS columns The flow-through fractions of the GST-TFIIS chromatography, which contained the majority of MCM proteins, were subsequently chromatographed on histone H3/H4-agarose

as described [38–40] The binding of different MCM2 deletion mutants to GST-TFIIS-Sepharose or H3/H4 histones relative to endogenous MCM2 and other MCM proteins was analyzed by Western Blotting

Plasmids encoding for recombinant MCM2 polypeptides (Fig 1) were transfected into 293 cells and expression was

Fig 1 Scheme of constructs for the expression of FLAG-tagged MCM2 deletion mutants Human MCM2 DNA sequences encoding the indicated amino acid residues in the full length protein were cloned into pFLAG-CMV-2 All sequences contain a N-terminal FLAG tag (+) denotes that binding of the expressed protein to pol II holoenzyme

or histone H3/H4 was observed (–) denotes that no binding was observed (–/+) denotes very weak binding.

allowed to proceed for 36–48 h Whole cell extracts were

prepared as described previously [6] Under these conditions

all recombinant MCM2 polypeptides were expressed and

extracted at levels, which were roughly equal to the

endogenous MCM2 with the exception of MCM2(198–

892), which was expressed at somewhat lower levels (Fig 2

and data not shown)

Binding of MCM deletion mutants to GST-TFIIS beads

GST-TFIIS affinity chromatography was used to purify

RNA pol II holoenzyme and associated MCM proteins

from whole cell extracts Each extract was passed through a

control column containing GST alone and then loaded on a

GST-TFIIS column Both columns were washed extensively

with low-salt buffer and eluted with 0.325MNaCl and then

with 1M NaCl Western blots of the load, flow-through,

wash, 0.325Meluate, and 1Meluate fractions are shown in

Fig 3 As reported previously [5,6], pol II bound to

GST-TFIIS and eluted as two distinct fractions at 0.325MNaCl,

which corresponds to pol II holoenzyme, and 1M NaCl,

which corresponds to core pol II (Fig 3, lanes 1–9) The

amounts of pol II in the 0.325MNaCl eluates were similar

between different extracts as detected by an antibody

against its largest subunit [5,6] There were noticeable

differences in the intensity of pol II signal in the 1Meluate

probably because of the limiting quantities of antigen in this

fraction that occasionally might be below the sensitivity of our antibody Importantly, in the 0.325MNaCl eluate from all extracts we detected comparable amounts of endogenous

Fig 2 Expression of FLAG-tagged MCM2 deletion mutants in 293 cells Plasmids encoding N-terminally FLAG-tagged MCM2 deletion mutants were individually transfected in 293 cells and expressed for 36–48 h Whole cell extracts were prepared as described in Materials and methods and 90 lg were loaded per lane Proteins were separated

on 10% SDS/PAGE gels and analyzed by Western blotting with anti-FLAG Ig.

Fig 3 Analysis of binding of FLAG-tagged MCM2 deletion mutants to GST-TFIIS Affinity columns (100 lL) containing GST (10 mgÆmL)1) or GST-TFIIS (10 mgÆmL)1) were loaded in series with 20–40 mg of 293 whole-cell extract, washed extensively and eluted with 0.325 M NaCl and then with 1 M NaCl 0.33% of the Load (L), flowthrough (FT), and 33% of the final wash (W) and eluate (E) fractions were analyzed by Western blotting with the indicated antibodies The figure shows one of three independent experiments (lines a–d) or one of two independent experiments (lines e–i).

MCM2 (Fig 3, lanes 10–18) Binding of full length

FLAG-tagged MCM2 to GST-TFIIS is shown in Fig 3, row (a),

lanes 19–27

Western blotting with anti-FLAG Ig was used to

compare the binding of MCM2 deletion mutants to

GST-TFIIS relative to the endogenous MCM2 (Fig 3, lanes 10–

18) and to the full length FLAG-tagged MCM2 Binding to

GST-TFIIS was considered positive when signals in the

0.325MNaCl eluates (Fig 3, lanes 16 and 26) were stronger

than the signals in the final wash fractions (Fig 3, lanes 15

and 25) While N-terminal deletions such as MCM2(198–

892) and MCM2(345–892) (Fig 3, rows g, h) displayed

deficient association with GST-TFIIS, the C-terminal

deletion mutants MCM2(1–230), MCM2(1–247), and

MCM2(1–511)(Fig 3, rows d–f ) all coeluted with pol II

holoenzyme at comparable levels to that of full

length-MCM2(1–892) (Fig 3, row a) However, further C-terminal

deletions MCM2(1–197) and MCM2(1–167) caused an

incremental decrease and disappearance of the FLAG

signal in the 0.325 M eluate (Fig 3, rows b, c), while

endogenous MCM2 signal in this fraction was similar for all

extracts These initial results suggested that amino acids

168–230 of MCM2 could be involved in the interaction

between MCM proteins and pol II holoenzyme

Binding of MCM deletion mutants to H3/H4 dimers

MCM proteins bind to histone H3/H4 dimers in vitro via an

interaction mediated by the N terminus of MCM2 [38–40]

It is possible that the interaction between MCMs and the

pol II holoenzyme is mediated by histones, which could be recruited to the holoenzyme in a specific or nonspecific manner We tested this possibility by analyzing the binding

of the MCM2 mutants to histone H3/H4 The flow-through fractions of the GST-TFIIS columns were loaded sequen-tially on BSA-agarose and H3/H4-agarose columns The resins were washed with low salt buffer and eluted with 0.5, 0.75, and 2M NaCl Load, flow-through, final wash and eluate fractions were analyzed by Western blotting First we examined the binding of MCMs to the H3/H4 resin using an antibody against the conserved ATP binding domain of all MCM proteins [49] This antibody cross-reacts with many bands in crude extracts; however, in purified fractions it detects three to five bands, which correspond to MCM proteins [36,39,51] In the wash fraction and the eluates of the H3/H4 columns we detected three to five bands with the expected mobility of MCM2, MCM3, MCM4, MCM5 and MCM6 (Fig 4, lanes 5–7) These signals were significantly higher than the correspond-ing background signals from the control BSA-agarose columns (Fig 4, lanes 1–4) We also detected recombinant MCM2(198–892) and MCM2(345–892), which contained intact ATP binding domain (data not shown) Because this antibody has a low affinity, we did not observe exactly the same profile of bands in all eluates possibly because some of the MCM proteins were present below the threshold of detection We did not see any MCMs in the 2MNaCl eluate

by this or other anti-MCM Ig (not shown) This observation

is in disagreement with the previously reported histone H3/ H4 chromatography experiments, in which the majority of

Fig 4 Analysis of binding of FLAG-tagged MCM2 deletion mutants to H3/H4-agarose Affinity columns (100 lL) containing bovine serum albumin (BSA, 5 mgÆmL)1) and histone H3/H4 (5 mgÆmL)1) were loaded in series with 15–30 mg of GST-TFIIS column flow-through, washed and eluted with buffer A containing 0.5 M NaCl, buffer A containing 0.75 M NaCl, and buffer A containing 2 M NaCl 0.33% of the Load (L), flowthrough (FT), wash (W) and eluate (E) fractions were analyzed by Western blotting with the indi-cated antibodies In rows a, e, f, g, and h, we pooled all wash fractions and eluates and loaded 33% of each per lane, respectively In rows b, c, and d, we show 33% of final wash and 33% of the pooled 0.5 M and 0.75 M eluate The figure shows one of three inde-pendent experiments (lines a–d) or one of two independent experiments (lines e–i).

MCMs were found in the 2Msalt eluate [38–40] We do not

understand this discrepancy Nonetheless, our histone

H3/H4 resin specifically retained MCM proteins as reported

in [38–40] and was considered adequate for further analyses

Next we examined the binding of the MCM2 deletion

mutants to H3/H4 relative to the endogenous MCM2 and

to the full length FLAG-tagged MCM2 Significant

amounts of endogenous MCM2 were found in all histone

H3/H4 eluates while no MCM2 was detected in the

corresponding eluates from the control BSA-agarose

col-umn (Fig 4, lanes 8–14) In agreement with previous

reports [38–40], the N-terminal deletion mutants

MCM2(198–892) and MCM2(345–892) did not associate

with histone H3/H4 (Fig 4, lanes 15–21, rows g and h) All

other recombinant polypeptides closely resembled the

elution pattern of the endogenous MCM2 (Fig 4, lanes

15–21) Importantly, the MCM2(1–167) and MCM2(1–

197), which did not bind to GST-TFIIS, bound strongly to

histones (Fig 4, lanes 15–21, rows b and c)

This second set of experiments clearly demonstrated that

the sequence of MCM2, which confers association with

pol II holoenzyme (amino acids 168–230, Fig 3) is distinct

from the sequence, which is required for its association with

histone H3/H4 (amino acids 1–167, Fig 4) [38–40]

Binding of pol II holoenzyme to GST-MCM2(169–212)

We tested the possibility that amino acids 168–230 of

MCM2 were required for binding to pol II holoenzyme by a

different procedure We expressed this peptide as a

GST-fusion protein and used it in affinity chromatography

experiments with HeLa cell extract Because the C-terminus

of 168–230 contains the peptide VNYEDLA, which is part

of the reported HBO1 binding site [41], we decided to

further truncate the C-terminus of this sequence to produce

GST-MCM2(169–212) fusion protein The protein was

coupled to glutathione beads and HeLa cell extract was

passed in parallel through GST, TFIIS and

GST-MCM2(169–212) beads, the beads were washed extensively

with CB buffer and eluted with 1M NaCl The load,

flowthrough, final wash and eluate fractions were analyzed

by Western blotting (Fig 5) In the eluates of both

GST-TFIIS and GST-MCM2(169–212) columns we detected the

largest subunit of pol II together with subunits of the

general transcription factors TFIID (TBP), TFIIH(p62)

and TFIIB Other components of pol II holoenzyme such

as CycC(SRB10) and CPSF(p160) were present at

signifi-cantly lower levels in the GST-MCM2(169–212) eluates

relative to the GST-TFIIS eluates A component of the

origin recognition complex (ORC2), which associates with

MCM proteins at origins of replication [52], was not

detected in the eluates of both columns indicating that the

observed signals are not a consequence of contamination by

extract or chromatin In all Western blots very little or no

signal from all antigens was observed in the final wash

fractions and control GST eluates

In summary, we observed that MCM2(169–212) was

binding pol II, TFIID, TFIIH and TFIIB with similar

(albeit lesser) efficiency as compared to a bona fide

holoenzyme binding ligand, TFIIS [5,6] This data is

consistent with the idea that amino acids 169–212 of

MCM2 are binding to some component(s) of RNA

polymerase II holoenzyme

Point mutations in the MCM2(169–212) domain disrupt the binding of pol II holoenzymein vitro and in vivo

If the MCM2(169–212) domain is required for binding to pol II holoenzyme, then mutations in this region should disrupt the interactions, which were described in Figs 3 and

5 To test this hypothesis, we substituted conserved hydro-philic and hydrophobic residues in GST-MCM2(169–212)

as shown in Fig 6A GST, GST-MCM2(169–212)V203E/ F204L, MCM2(169–212)L192P/R193G, and GST-MCM2(169–212)wt were coupled to beads at 10 mgÆmL)1 (data not shown) and assayed for their ability to pull down components of pol II holoenzyme from HeLa extract Each column was loaded with 100 mg of HeLa whole cell extract, washed extensively with CB, and eluted with 1M NaCl Flowthrough, final wash and eluate fractions were analyzed

by Western blotting with antibodies against pol II and several general transcription factors (Fig 6B) Consistent with our previous experiment (Fig 5), considerable amounts of pol II, TFIID(TBP), TFIIH(p62) and TFIIB were detected in the GST-MCM2(169–212)wt eluate (Fig 6B, lane 12), while little to no signal was detected in the wash fractions or in the GST control eluate (Fig 6B, lane 3) GST-MCM2(169–212)L192P/R193G retained sim-ilar or slightly lower amounts of pol II, TBP, TFIIH(p62) and TFIIB (Fig 6B, lane 9) relative to GST-MCM2(169– 212)wt A dramatic decrease in the signals from all peptides was observed in the eluate of the GST-MCM2(169– 212)V203E/F204L column (Fig 6B, lane 6) These results

Fig 5 GST-MCM2(169–212) chromatography Affinity columns (250 lL) containing GST, GST-MCM2(169–212), or GST-TFIIS (each at 10 mgÆmL)1) were loaded in parallel with 50 mg of HeLa cell extract, washed extensively and eluted with 1 M NaCl 0.33% of the Load, flow-through, and 40% of the final wash and eluate fractions were analyzed by Western blotting with the indicated antibodies The figure shows one of two independent experiments.

clearly indicated that the V203E/F204L substitution

dis-rupted the ability of the MCM2(169–212) peptide to bind

pol II and general transcription factors in vitro

Next we tested if the same mutations would interfere with

the binding of MCM2 to pol II holoenzyme in vivo We

introduced the V203E/F204L and L192P/R193G

substitu-tions into the pMCM2(1–230) plasmid

FLAG-MCM2(1–230) was the smallest stable deletion mutant,

which contained the 169–212 region and retained full

capacity to copurify with pol II holoenzyme (Fig 3) The

wild type and mutant proteins were expressed in 293 cells

and extracted at comparable levels (Fig 7A) Each extract

was assayed by GST-TFIIS affinity chromatography Load,

flowthrough, wash and 0.325MNaCl eluate fractions were

analyzed by Western blotting with CTD Ig, and

anti-MCM2 Ig, which recognized equally well the endogenous

MCM2 and all MCM2(1–230) recombinant peptides

(Fig 7A) As observed in our previous experiment (Fig 3)

the largest subunit of pol II and endogenous MCM2

coeluted in the 0.325M NaCl eluate of the GST-TFIIS column The amounts of these two proteins in this fraction were similar for all extracts (Fig 7B, lanes 7 and 14) We then compared the copurification of each MCM2(1–230) mutant relative to endogenous MCM2 and to the MCM2(1–230)wt control (Fig 7B, row a) MCM2(1– 230)L192P/R193G (Fig 7B, row c, lane 14) displayed somewhat deficient association with GST-TFIIS, whereas MCM2(1–230)V203E/F203L was almost completely absent from the pol II holoenzyme fraction (Fig 7B, row b, lane 14) The relative abundance of the wild type and mutant MCM2(1–230) peptides in the eluates from the GST-TFIIS columns was confirmed by Western blot with anti-FLAG antibodies (data not shown) These results are in agreement with the results obtained with GST-fusion proteins in vitro (Fig 6) We conclude that the V203E/F203L substitution disrupts the association of MCM2 with pol II holoenzyme both in vitro and in vivo The experiments presented in Figs 6 and 7 further verify the importance of amino acids 169–212 to the interaction between MCM2 and RNA pol II holoenzyme

D I S C U S S I O N

MCM2 interacts with pol II holoenzyme

We have previously shown that several MCM proteins copurified with pol II holoenzyme preparations from human and Xenopus cells [6] In this paper we show that

a likely site, which mediates this interaction is amino acids 169–212 of MCM2 Our results demonstrate that in vivo expressed epitope-tagged MCM2 deletion mutants bind to GST-TFIIS and elute in the pol II holoenzyme fraction

as shown before [5,6] only if they contain amino acids 168–230 (Fig 3) N- and C-terminal deletions within this region significantly decreased, but did not completely abolish binding to GST-TFIIS (Fig 3, lines c and g) The C-terminal deletions MCM2(1–167) and MCM2(1– 197) retained their ability to associate with histones H3/ H4 (Fig 4, lines b and c) as previously reported [38–40] Therefore, it is unlikely that their deficiency in binding pol II holoenzyme is a result of misfolding or inactivation

We do not have a reliable assay to test the functional status of MCM2(198–892) and MCM2(345–892) (Figs 3 and 4) and other N-terminal deletions beyond amino acid

345 (not shown), which did not bind to GST-TFIIS or histone H3/H4

In a separate set of experiments we show that a GST-MCM2(169–212) ligand binds pol II and several previously characterized components of RNA polymerase II holoen-zyme in vitro with similar efficiency as compared to GST-TFIIS (Fig 5) These experiments imply that MCM2 peptides might be recruited to the holoenzyme independ-ently of whether they are in a complex with other MCM proteins or not Furthermore, a previous study [53] indica-ted that the interaction between MCM2 and MCM4,6,7 is located in the C-terminus of MCM2 Previously

MCM2,4,6,7 [51,54,55] consist of single molecules of the six or four MCM proteins, respectively It is therefore unlikely that the deletion mutants were recruited to GST– TFIIS via interactions with complexes that already contain endogenous MCM2

Fig 6 Point mutations in the MCM2(169–212) domain interfere with

the binding of pol II and general transcription factors in vitro (A) The

indicated amino acid substitutions were introduced into the

GST-MCM2(169–212) peptide by site-directed mutagenesis (B) Affinity

columns (250 lL) containing GST, GST-MCM2(169–212)V203E/

F204L, GST-MCM2(169–212)L192P/R193G, or GST-MCM2(169–

212)wt (each at 10 mgÆmL)1) were loaded with 100 mg of HeLa whole

cell extract, washed extensively and eluted with 1 M NaCl 0.33% of the

Flowthrough (FT) and 40% of each final wash (W) and eluate (E)

fractions were analyzed by Western blotting with the indicated

anti-bodies The figure shows one of two independent experiments.

The idea that MCM2(169–212) is a site for interaction

with pol II holoenzyme is significantly strengthened by the

observation that a double amino acid substitution (V203E/

F204L) within this region disrupted the association of this

domain with pol II holoenzyme both in vitro and in vivo

(Figs 6 and 7) Taken together, our results strongly suggest

that amino acids 169–212, which are not required for the

interaction of MCM2 with histones H3/H4 [39,40] (Fig 4)

or HBO1 [41], are involved in the interaction with some

component(s) of pol II holoenzyme We propose that three

different sites of MCM2 are involved in independent

interactions with histone H3/H4, pol II holoenzyme and

HBO1 as shown in Fig 8

In light of our previous study [6] the most likely

explanation of our observations is that MCM2(169–212)

associates with some component(s) of pol II holoenzyme

directly or via a bridging interaction The identity of this

component is not known We have shown that

antibod-ies against the carboxyterminal domain of the largest

subunit of pol II (CTD) disrupt the interaction between

pol II holoenzyme and MCM proteins [6] In addition,

MCMs, CPSF, CstF, and TFIIH bound to recombinant

CTD [6] It is possible that the partner of the

MCM2(169–212) peptide in pol II holoenzyme is within

one of these three factors or it is the CTD itself,

however, other components of the pol II holoenzyme

should also be considered

Characteristics of the MCM2(169–212) sequence The MCM2(169–212) sequence displays no obvious fea-tures that might suggest possible function (See Fig 8) It has

a pI of 9.18 and evenly distributed positively and negatively charged residues Jpred2 indicates two stretches of potential helices, however, no homology to previously characterized protein–protein interaction domains and no obvious simi-larity with known 3D structures as determined by SWISS -MODELand 3D-PSSMwere detected These analyses give us little hint about the nature of this domain and the possible mechanism of interaction with potential target peptides NETPHOS2.0 indicated three high score phosphorylation sites

in the putative pol II holoenzyme binding sequence of MCM2 (Fig 8) It is conceivable that interaction between MCMs and holoenzyme is regulated by phosphorylation at the predicted serine residues, but the significance of these residues is yet to be established

The 169–212 amino acid sequence of human MCM2 has highly homologous counterparts in mouse, Xenopus, and Drosophila (Fig 8), suggesting that a similar interaction between MCM2 and pol II holoenzyme might be taking place in these organisms The homology with the cognate MCM2 regions in S pombe, C elegans and S cerevisiae is

56, 54 and 46%, respectively, with several highly conserved hydrophobic residues, which are identical in all species (Fig 8) The sequence similarities do not provide clues as to

Fig 7 Point mutations in the MCM2(169–212) domain diminish binding of pol II holoenzyme in vivo (A) The amino acid substitutions, shown in Fig 6A, were introduced into the pFLAG-MCM2(1–230) plasmid by site directed mutagenesis N-Terminally flag-tagged full-length MCM2(1– 892) (lanes 1 and 5), MCM2(1–230)V203E/F204L (lanes 2 and 6), MCM2(1–230)L192P/R193G (lanes 3 and 7), and MCM2(1–230)wt (lanes 4 and 8) were expressed in 293 cells and whole cell extracts were prepared and 110 lg were loaded per lane Proteins were resolved on 10% SDS/PAGE gels and analyzed by Western blotting with anti-FLAG (lanes 1–4) and anti-MCM2 (lanes 5–8) Ig eMCM2, endogenous MCM2 (B) GST-TFIIS affinity chromatography was conducted as described in Materials and methods 0.33% of the Load (L), flowthrough (FT), and 50% of the final wash (W) and 0.325 M NaCl eluate (E) fractions were analyzed by Western blotting with antipol II CTD and anti-MCM2 Ig eMCM2, endogenous MCM2 The relative net intensities of bands in Lane 14 are plotted and the position of each MCM2(1–230) mutant and its relative eMCM2 control are indicated The figure shows one of three independent experiments.

whether MCM2 and pol II holoenzyme bind to each other

in these organisms In this context it is important that MCM

proteins were not detected in any of the S cerevisiae pol II

holoenzyme or mediator complexes despite extensive

bio-chemical analyses [2,8,14,56]

Functional significance of the pol II holoenzyme–MCM

interaction

The functional importance of the established interaction

between MCM proteins and pol II holoenzyme remains

enigmatic MCM proteins are components of the DNA

replication licensing machinery [24,35] Their association

with pol II holoenzyme could reflect some function of the

latter complex in replication Several findings are in tune

with this idea Recruitment of pol II holoenzyme to

origins of DNA replication in S cerevisiae significantly

stimulates their activity [45] In addition, transcriptional

activators, most of which interact with pol II holoenzyme,

stimulate DNA replication when tethered to v iral or

cellular origins [57–62] We found a genetic interaction

between pol II CTD and mcm5 and also showed

associ-ation of pol II with origins of DNA replicassoci-ation in

S cerevisiae (K Yankulov, D Kramer & R Dziak.,

unpublished observations) Control of mammalian origins

of replication is less well understood, yet it has been

reported that mammalian pol II holoenzyme complexes

contains proteins, which function in DNA repair and

replication [3,19] Furthermore, some mammalian origins

of DNA replication have been found within mammalian

enhancers [63,64], which presumably interact with pol II

holoenzyme

Another possibility is that MCM proteins function in

pol II transcription For example, stimulation of IFN-c

responsive genes is significantly decreased by mutations in

Stat1a, which also preclude the association of this protein

with MCM3/MCM5 complex [43,44] We showed that

antibodies against MCM2 inhibit pol II transcription in

injected Xenopus oocytes Effects on DNA replication were not analyzed because these cells do not replicate DNA [6]

We also found some transcriptional deficiencies in mcm5 mutants in S cerevisiae (K Yankulovand D Leishman, unpublished results) Preliminary experiments also indicate consistent inhibition of the expression of a plasmid-borne reporter gene upon overexpression of MCM2 with muta-tions in the holoenzyme interacting domain (data not shown)

It seems that the observed association between MCM proteins and pol II holoenzyme could reflect some addi-tional roles of these two complexes in pol II transcription and DNA replication, respectively At present, the mech-anism of action that signifies this association is unclear It is conceivable that histones, pol II holoenzyme, and ORC could come in close proximity via contacts with MCM2 to positively (or negatively) regulate origin function MCM2 may also have some unknown role in mediating pol II– histone contacts Clearly, addressing these questions requires an in vivo system where effects on DNA replication, transcription, and the cell cycle can be analyzed The identification of mutations in MCM2, which preclude its interaction with pol II holoenzyme, is therefore a major step towards such a detailed analysis MCM2(169–212) peptides

or full length MCM2 with mutations within the pol II holoenzyme binding region can be tested for dominant negative effects in stably transfected cells This approach will provide opportunities for a focused functional analysis

of the MCM–pol II holoenzyme interaction

A C K N O W L E D G E M E N T S

We thank N Thompson and I Todorovfor gifts of antibodies;

I Todorovfor the pBS/BM28 plasmid; R Lu for help with generation

of antibodies; R Mosser, A Wildeman, D Evans, J Bag, G Harauz,

D Leishman for valuable suggestions and discussion This study was supported by a grant from Canadian Institutes for Health Research (CIHR no 36371) to K Yankulov.

Fig 8 Characteristics of the MCM2(169–212) peptide The relative positions of the histone H3, pol II holoenzyme, HBO1 binding sites and the ATP homology domain in MCM2 are shown (not to scale) Similarity search and multiple sequence alignment to the approximate human pol II holoenzyme-binding sequence were performed by BLAST Amino acid residues identical to the human ones are represented by (.) Overall percentage homology to the human peptide is shown under the name of the species (*) indicates a potential site of phosphorylation The solid bar above the human sequence indicates predicted helices in the peptide Bolded text highlights the amino acids that we substituted for in our point mutation analyses.

R E F E R E N C E S

1 Orphanides, G., Lagrange, T & Reinberg, D (1996) The general

transcription factors of RNA polymerase II Genes Dev 10, 2657–

2683.

2 Koleske, A.J & Young, R.A (1994) An RNA polymerase-II

holoenzyme responsive to activators Nature 368, 466–469.

3 Maldonado, E., Shiekhattar, R., Sheldon, M., Cho, H., Drapkin,

R., Rickert, P., Lees, E., Anderson, C., Linn, S & Reinberg, D.

(1996) A human RNA polymerase II associated complex with

SRB and DNA-repair proteins Nature 381, 86–89.

4 Ossipow, V., Tassan, J.P., Nigg, E.A & Schibler, U (1995) A

mammalian RNA polymerase II holoenzyme containing all

components required for promoter-specific transcription

initia-tion Cell 83, 137–146.

5 Pan, G., Aso, T & Greenblatt, J (1997) Interaction of elongation

factors TFIIS and elongin A with a human RNA polymerase II

holoenzyme capable of promoter-specific initiation and responsive

to transcriptional activators J Biol Chem 272, 24563–24571.

6 Yankulov, K., Todorov, I., Romanowski, P., Licatalosi, D., Cilli,

K., McCracken, S., Laskey, R & Bentley, D.L (1999) MCM

proteins are associated with RNA polymerase II holoenzyme Mol

Cell Biol 19, 6154–6163.

7 Myers, L.C.G.C., Bushnell, D.A., Lui, M., Erdjument-Bromage,

H., Tempst, P & Kornberg, R.D (1998) The Medical proteins of

yeast and their function through the RNA polymerase II

carboxy-terminal domain Genes Dev 12, 45–54.

8 Kim, Y.J., Bjorklund, S., Li, Y., Sayre, M.H & Kornberg, R.D.

(1994) A multiprotein mediator of transcriptional activation and

its interaction with the C-terminal repeat domain of RNA

poly-merase II Cell 77, 599–608.

9 Hengartner, C.J., Thompson, C.M., Zhang, J., Chao, D.M., Liao,

S.M., Koleske, A.J., Okamura, S & Young, R.A (1995)

Asso-ciation of an activator with an RNA polymerase II holoenzyme.

Genes Dev 9, 897–910.

10 Gu, W., Malik, S., Ito, M., Yuan, C.X., Fondell, J.D., Zhang, X.,

Martinez, E., Qin, J & Roeder, R.G (1999) A novel human SRB/

MED-containing cofactor complex, SMCC, involved in

tran-scription regulation Mol Cell 3, 97–108.

11 Sun, X., Zhang, Y., Cho, H., Rickert, P., Lees, E., Lane, W &

Reinberg, D (1998) NAT, a human complex containing Srb

polypeptides that functions as a negative regulator of activated

transcription Mol Cell 2, 213–222.

12 Thompson, C.M & Young, R.A (1995) General requirement for

RNA polymerase II holoenzymes in vivo Proc Natl Acad Sci.

USA 92, 4587–4590.

13 Li, Y., Bjorklund, S., Jiang, Y.W., Kim, Y.J., Lane, W.S.,

Still-man, D.J & Kornberg, R.D (1995) Yeast global transcriptional

regulators Sin4 and Rgr1 are components of mediator complex/

RNA polymerase II holoenzyme Proc Natl Acad Sci USA 92,

10864–10868.

14 Wilson, C.J., Chao, D.M., Imbalzano, A.N., Schnitzler, G.R.,

Kingston, R.E & Young, R.A (1996) RNA polymerase II

holoenzyme contains SWI/SNF regulators involved in chromatin

remodeling Cell 84, 235–244.

15 Han, S.J.L.Y., Gim, B.S., Ryu, G.H., Park, S.J., Lane, W.S &

Kim, Y.J (1999) Activator-specific requirement of yeast mediator

proteins for RNA polymerase II transcriptional activation Mol.

Cell Biol 19, 979–988.

16 Gustafsson, C.M., Myers, L.C., Li, Y., Redd, M.J., Lui, M.,

Erdjument, B.-H., Tempst, P & Kornberg, R.D (1997)

Identifi-cation of Rox3 as a component of mediator and RNA polymerase

II holoenzyme J Biol Chem 272, 48–50.

17 Hampsey, M (1998) Molecular genetics of the RNA polymerase

II general transcriptional machinery Microbiol Mol Biol Rev 62,

465–503.

18 Carlson, M (1997) Genetics of transcriptional regulation in yeast: connections to the RNA polymerase II CTD Annu Rev Cell Dev Biol 13, 1–23.

19 Scully, R., Anderson, S.F., Chao, D.M., Wei, W.YeL., Young, R.A., Livingston, D.M & Parvin, J.D (1997) BRCA1 is a com-ponent of the RNA polymerase II holoenzyme Proc Nat Acad Sci USA 94, 5605–5610.

20 Anderson, S.F., Schlegel, B.P., Nakajima, T., Wolpin, E.S & Parvin, J.D (1998) BRCA1 protein is linked to the RNA poly-merase II holoenzyme complex via RNA helicase A Nat Genet.

19, 254–256.

21 McCracken, S., Fong, N., Yankulov, K., Ballantyne, S., Pan, G.H., Greenblatt, J., Patterson, S.D., Wickens, M & Bentley, D.L (1997) The C-terminal domain of RNA polymerase II cou-ples messenger-RNA processing to transcription Nature 385, 357– 361.

22 Chang, M & J.J (1997) A multiplicity of mediators: alternative forms of transcription complexes communicate with transcrip-tional regulators Nucleic Acids Res 25, 4861–4865.

23 Kearsey, S.E & L.K (1998) MCM proteins: evolution, properties, and role in DNA replication Biochim Biophys Acta 1398, 113– 136.

24 Lei, M & Tye, B.K (2001) Initiating DNA synthesis: from recruiting to activating the MCM complex J Cell Sci 114, 1447– 1454.

25 Newlon, C.S (1997) Putting it all together: building a prereplica-tive complex Cell 91, 717–720.

26 Takisawa, H., Mimura, S & Kubota, Y (2000) Eukaryotic DNA replication: from pre-replication complex to initiation complex Curr Opin Cell Biol 12, 690–696.

27 Blow, J.J & Laskey, R.A (1988) A role for the nuclear envelope in controlling DNA replication within the cell cycle Nature 332, 546– 548.

28 Aparicio, O.M., Weinstein, D.M & Bell, S.P (1997) Components and dynamics of DNA replication complexes in S cerevisiae: redistribution of MCM proteins and Cdc45p during S phase Cell.

91, 59–69.

29 Coue, M., Kearsey, S & M.Mechali (1996) Chromatin binding, nuclear localization and phosphorylation of Xenopus cdc21 are cell-cycle dependent and associated with the control of initiation of DNA replication EMBO J 15, 1085–1097.

30 Liang, C & S.B (1997) Persistent initiation of DNA replication and chromatin-bound MCM proteins during the cell cycle in cdc6 mutants Genes Dev 11, 3375–3386.

31 Tanaka, T., Knapp, D & Nasmyth, K (1997) Loading of an Mcm protein onto DNA replication origins is regulated by Cdc6p and CDKs Cell 90, 649–660.

32 Todorov, I.T., Attaran, A & Kearsey, S.E (1995) BM28, a human member of the MCM2-3–5 family, is displaced from chromatin during DNA replication J Cell Biol 129, 1433–1445.

33 Chen, Y., Hennessy, K.M., Botstein, D & Tye, B.K (1992) CDC46/MCM5, a yeast protein whose subcellular localization is cell cycle-regulated, is involved in DNA replication at autono-mously replicating sequences Proc Natl Acad Sci USA 89, 10459–10463.

34 Labib, K., Tercero, J.A & Diffley, J.F (2000) Uninterrupted MCM2-7 function required for DNA replication fork progression Science 288, 1643–1647.

35 Labib, K & Diffley, J.F (2001) Is the MCM2-7 complex the eukaryotic DNA replication fork helicase? Curr Opin Genet Dev.

11, 64–70.

36 Ishimi, Y (1997) A DNA helicase activity is associated with an MCM4-6 and -7 complex J Biol Chem 272, 24508–24513.

37 Ishimi, Y & Komamura-Kohno, Y (2001) Phosphorylation of Mcm4 at specific sites by cyclin-dependent kinase leads to loss of Mcm4,6,7 helicase activity J Biol Chem 13, 13.