Multiple Interactions of the Intrinsically Disordered Region between ...

Multiple Interactions of the Intrinsically Disordered Regionbetween the Helicase and Nuclease Domains of the ArchaealHef Protein*□S

Received for publication, January 31, 2014, and in revised form, June 18, 2014 Published, JBC Papers in Press, June 19, 2014, DOI 10.1074/jbc.M114.554998

Sonoko Ishino‡, Takeshi Yamagami‡, Makoto Kitamura‡, Noriyuki Kodera§¶, Tetsuya Mori§¶, Shyogo Sugiyama§¶,Toshio Ando§¶, Natsuko Goda�, Takeshi Tenno�, Hidekazu Hiroaki�, and Yoshizumi Ishino‡1

From the ‡Department of Bioscience and Biotechnology, Graduate School of Bioresource and Bioenvironmental Sciences, andFaculty of Agriculture, Kyushu University, Fukuoka 812-8581, the §Bio-AFM Frontier Research Center and ¶Department of Physics,College of Science and Engineering, Kanazawa University, Kanazawa 920-1192, and the �Graduate School of PharmaceuticalSciences, Nagoya University, Nagoya 464-8601, Japan

Background: Hef/FANCM participates in interstrand cross-link DNA repair.Results: The predicted intrinsically disordered region (IDR) in Hef was experimentally verified. Proliferating cell nuclearantigen and a RecJ-like protein interact with the IDR individually, but not simultaneously.Conclusion: The IDR in Hef interacts with multiple proteins.Significance: Hef may function in DNA repair by association of its IDR with multiple partners.

Hef is an archaeal protein that probably functions mainly installed replication fork repair. The presence of an unstructuredregion was predicted between the two distinct domains of theHef protein. We analyzed the interdomain region of Thermococ-cus kodakarensis Hef and demonstrated its disordered structureby CD, NMR, and high speed atomic force microscopy (AFM).To investigate the functions of this intrinsically disorderedregion (IDR), we screened for proteins interacting with the IDRof Hef by a yeast two-hybrid method, and 10 candidate proteinswere obtained. We found that PCNA1 and a RecJ-like proteinspecifically bind to the IDR in vitro. These results suggested thatthe Hef protein interacts with several different proteins thatwork together in the pathways downstream from stalled repli-cation fork repair by converting the IDR structure depending onthe partner protein.

Hef is an archaeal protein that most probably functionsmainly in stalled replication fork repair. It was originallydiscovered by screening for factors associating with thearchaeal Holliday junction resolvase, Hjc, from a total cellextract of Pyrococcus furiosus, a hyperthermophilic archaeon(1). Identification of the gene corresponding to this activityrevealed that the encoded protein consists of two distinctdomains, with similarity to the DEAH helicase family and theXPF/Mus81 nuclease superfamily, respectively (1). Biochemi-cal characterization of the purified protein confirmed the spe-cific affinity for branched DNA structures, including the repli-cation fork. The N-terminal domain possesses ATPase andhelicase activities that were dramatically stimulated by fork-

structured DNAs, and the C-terminal domain has an endonu-clease activity that specifically cleaves nicked, flapped, and fork-structured DNAs (1, 2). Based on these properties, this proteinwas designated as Hef (helicase-associated endonuclease forfork-structured DNA), and it was proposed that Hef may be aprototypical enzyme for the eukaryotic XPF/Rad1/Mus81nuclease family (2, 3).

A genetic manipulation system was developed for Thermo-coccus kodakarensis (4), and phenotype analyses of the hefmutant strain confirmed that Hef is involved in multiple repairprocesses, and its especially high sensitivity to mitomycin Csuggested that Hef performs a critical function in DNA inter-strand cross-link repair (5). The helicase and nuclease activitiesof the purified T. kodakarensis Hef protein (TkoHef) for fork-structured DNA was also confirmed in vitro (5). These geneticand biochemical properties of TkoHef suggested that this pro-tein actually works at stalled replication forks in T. kodakarensisby the coordination of its helicase and endonuclease activities,as predicted.

Hef is well conserved in Euryarchaeota. Genetic and cytolog-ical analyses of the hef gene in Haloferax volcanii also revealedthat Hef is involved in stalled replication fork repair (6, 7). Onthe other hand, the crenarchaeal organisms have a proteincomprising only the endonuclease domain, and its nucleaseactivity for branched DNA is completely proliferating cellnuclear antigen (PCNA)2-dependent (8, 9).

Research on stalled replication fork repair in Eukarya is nowactively progressing (10 –13). Defects in this repair system leadto diseases with symptoms including developmental abnormal-ities and a predisposition to cancer. A well known hereditarydisease related to this repair system is Fanconi anemia (FA).

* This work was supported by Ministry of Education, Culture, Sports, Scienceand Technology of Japan Grants 21113005, 23310152, and 26242075 (toY. I.) and 21113002 (to T. A.).

□S This article contains supplemental Movies S1–S3.1 To whom correspondence should be addressed: 6-10-1 Hakozaki, Higashi-

ku, Fukuoka, Fukuoka 812-8581, Japan. Tel.: 81-92-642-4217; Fax: 81-92-642-3085; E-mail: [email protected].

2 The abbreviations used are: PCNA, proliferating cell nuclear antigen; IDP,intrinsically disordered protein; IDR, intrinsically disordered region; HSQC,heteronuclear single quantum coherence; AFM, atomic force microscopy;SPR, surface plasmon resonance; TEV, tobacco etch virus; FA, Fanconi ane-mia; Y2H, yeast two-hybrid; HAN, Hef-associated nuclease.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 289, NO. 31, pp. 21627–21639, August 1, 2014© 2014 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

AUGUST 1, 2014 • VOLUME 289 • NUMBER 31 JOURNAL OF BIOLOGICAL CHEMISTRY 21627

by guest on February 6, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Currently, 16 genes, designated as FANCA to FANCP, havebeen identified in the FA-related DNA repair pathway (14 –18).Among these gene products, the FANCM protein works as amolecular scaffold with FAAP24 and MHF proteins to recruitthe FA core complex, containing FANCA, -G, -C, -E, -F, -B, -L,and FAAP100, for the monoubiquitination of FANCD2 andFANCI. These modifications activate the FANCD2-I complexto recruit other related proteins, including BRCA1, BRCA2,FANCN, and FANCJ, to the damaged site. Reports in 2005 thathuman and chicken FANCMs are orthologs of the archaeal Hef(19, 20) stimulated progress of the research on the molecularmechanisms of FA repair pathways.

Comparisons of the structures of the archaeal Hef andeukaryotic FANCM proteins led us to focus on the function ofthe region between the N-terminal helicase and C-terminalnuclease, because the region seemed to be intrinsically disor-dered. Intrinsically disordered proteins (IDPs), which lack a sta-ble three-dimensional structure, are now attracting attention inboth protein chemistry and molecular biology fields, becausethe structural flexibility of IDPs confers functional advantages(21–24). The structural plasticity enables the proteins to bindstructurally distinct target molecules by structural conversion.The IDPs are also able to provide a larger surface as an interface,to interact with other molecules, and to escape steric hin-drances. These properties are advantageous for interactionswith numerous partners with high specificity and low affin-ity. Therefore, the IDPs function in many life processes byinteracting with several partner molecules in a time-depen-dent manner.

In this study, we performed structural and functional analy-ses of the intrinsically disordered region (IDR) in T. kodakaren-sis Hef. The predicted IDR was experimentally confirmed to beunstructured. The interaction of IDR with several proteins,including PCNA and a RecJ-like protein, suggest that Hef usesits IDR for sequential bindings of several proteins to process therepair pathway as a scaffold at the stalled fork.

EXPERIMENTAL PROCEDURES

Prediction of IDR in the Hef Family Proteins—To predict theIDR in TkoHef and human FANCM, the Hef/FANCM proteinswere subjected to a GTOP database (25) search and a DICHOTprediction analysis (26), which divide proteins into structuraldomains and IDR with high accuracy.

Production and Purification of TkoHef—Cloning of the hefgene and preparation of the TkoHef protein were described in

detail in our previous report (5). For the preparation of theHis-tagged IDR (His-ID), the gene fragment encoding Met490–Lys593 of TkoHef (Hef-ID) was amplified by PCR from T. koda-karensis DNA, using the primer set ID-F/ID-R including NdeIand NotI recognition sequences at their 5�-regions (shown inTable 1) for insertion into the NdeI-NotI sites of a modifiedpET-28a(�) vector (Novagen). The modification of pET-28a(�) was the conversion of the thrombin recognition se-quence to the TEV protease recognition sequence. The resul-tant plasmid was designated as pET28a(TEV)-ID. Escherichiacoli BL21-CodonPlus (DE3)-RIL cells (Agilent) harboringpET28a(TEV)-ID were grown at 37 °C in 1 liter of LB medium,containing 50 �g/ml of kanamycin and 34 �g/ml of chloram-phenicol. When the E. coli culture reached an A600 of 0.3,expression of the gene encoding His-ID was induced by addingisopropyl �-D-thiogalactopyranoside to 1 mM. After cultivationat 25 °C for 16 h, the cells were harvested by centrifugation, andthe cell lysate was prepared by sonication in 40 ml of buffer A(50 mM sodium phosphate, pH 7.8, 0.5 M NaCl, and 1 mM

PMSF). After centrifugation for 20 min at 23,700 � g, the super-natant was applied to a His-affinity column (HisTrap HP, GEHealthcare), which was eluted with an imidazole gradient from10 mM to 0.5 M. The fractions containing His-ID were dialyzedagainst buffer B (50 mM Tris-HCl, pH 8.0, 0.5 mM DTT, 0.1 mM

EDTA) including 0.1 M NaCl, and then subjected to cationexchange (HiTrap SP HP, GE Healthcare) chromatography andgel filtration (Superdex 200, GE Healthcare) chromatography.

Preparation of Mutant TkoHefs—For the construction of theIDR deletion mutant (TkoHef�ID), four primers were designedto fuse the two fragments via PCR. These primer sequences arelisted in Table 1. The DNAs encoding the N-terminal (1,476nucleotides) and C-terminal (669 nucleotides) fragments wereamplified individually, using primer sets Hef-F/del-R and del-F/Hef-R, respectively. A second PCR was performed to com-bine the N-terminal and C-terminal fragments, using a mixtureof the amplified fragments. A third PCR was performed usingthe amplified fragment mixture as the DNA templates, withprimer set Hef-F/Hef-R. The fragment was excised by NdeI-NotI digestion, and was inserted into the corresponding sites ofpET-21a(�) (Novagen). The resultant plasmid was designatedas pET21a-Hef-del. Amino acid substitutions for TkoHef wereintroduced into the hef gene by PCR-mediated mutagenesis(QuikChange site-directed mutagenesis kit; Stratagene), bysequentially using primers HefT695D-F/HefT695D-R and

TABLE 1Primers and oligonucleotides used in this study

Name Sequence

Hef-F 5�-dGGCATATGCAGTACCTCAGACGAGACCTCATAG-3�Hef-R 5�-dGGGCGGCCGCTTAAGCCTTATCCTCTTCCTCAACG-3�del-F 5�-dGAAAGGAGAAGGGGATGTTCGATGTGAGAAAGCCGAAGGGC-3�del-R 5�-dGCCCTTCGGCTTTCTCACATCGAACATCCCCTTCTCCTTTC-3�HefT695D-F 5�-dCGGGGAGCGATAGCGGCGGTGGACCTTGACTGGGGAGTTCCAATATTG-3�HefT695D-R 5�-dCAATATTGGAACTCCCCAGTCAAGGTCCACCGCCGCTATCGCTCCCCG-3�HefT800stop-F 5�-dGGGAGATAAGGAGGGTCATCTGAGCACCCTACGTTGAGGAAGAG-3�HefT800stop-R 5�-dCTCTTCCTCAACGTAGGGTGCTCAGATGACCCTCCTTATCTCCC-3�ID-F 5�-dCGCGCATATGTTCGATGCGATAAAGAAAATCG-3�ID-R 5�-dGGGCGGCCGCTTACTTCGGCTTTCTCACAAAGATGG-3�TK0155-F 5�-dGGGCATATGGTTGTGAAAGACTGTCCCG-3�TK0155-R 5�-dGGGGCGGCCGCTTAGGAAATCTTCTTCAGCTTCACG-3�

Intrinsically Disordered Region of Hef/FANCM

21628 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 289 • NUMBER 31 • AUGUST 1, 2014


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

HefT800stop-F/HefT800stop-R (Table 1), to make a TkoHefmutant (TkoHef-m) that cannot form a stable homodimer. Thepreparation of TkoHef�ID and TkoHef-m proteins was per-formed according to the same procedure used for the TkoHefprotein. The purified fractions were concentrated and stored at4 °C.

Preparation of TK0155 (HAN)—The genes for TK0155 wereamplified by PCR from T. kodakarensis DNA, using the primerset including NdeI and NotI recognition sequences at their5�-regions (shown in Table 1) for insertion into the NdeI-NotIsites of pET-21a(�). The resultant plasmid was introduced intoE. coli BL21-CodonPlus (DE3)-RIL cells and grown at 37 °C in 1liter of LB medium, containing 50 �g/ml of ampicillin and 34�g/ml of chloramphenicol. When the E. coli culture reached anA600 of 0.3, expression of the genes were induced by addingisopropyl �-D-thiogalactopyranoside to 1 mM. After cultivationat 25 °C for 16 h, the cells were harvested, and the cell lysate wasprepared by sonication in 40 ml of buffer B including 1 M NaCl.The soluble cell extracts were heated at 80 °C for 20 min. Theheat-resistant fraction was subjected to treatment with 0.15%polyethyleneimine. The proteins in the supernatants were pre-cipitated by 80% saturated ammonium sulfate. The precipitatewas resuspended in buffer B containing 0.5 M ammonium sul-fate, and subjected to chromatography on a 5-ml HiTrap Phe-nyl HP column, which was developed with a linear gradient of0.5– 0 M ammonium sulfate. The eluted protein fractions weredialyzed against buffer B, and then loaded onto a 5-ml HiTrap QHP column (GE Healthcare), which was developed with a lineargradient of 0.1– 0.8 M NaCl.

Quantification of Purified Proteins—The protein concentra-tions of the purified fractions were determined by measuringthe absorbance at 280 nm, with theoretical extinction coeffi-cients of 57,760, 57,760, 56,270, 1,490, and 84,230 M�1 cm�1 forTkoHef, TkoHef�ID, TkoHef-m, His-ID, and HAN, respec-tively, based on tryptophan and tyrosine content, as describedpreviously (27).

Nuclease Assay—A substrate single-stranded DNA (5�-CGA-ACTGCCTGGAATCCTGACGAACTGTAG) was labeled atthe 3� terminus with FITC (Hokkaido System Science Co., Ltd.).TK0155 (HAN) (50 nM) was incubated with 100 nM single-stranded DNA in 20 �l of reaction buffer (25 mM Tris-HCl, pH8.0, 2 mM MgCl2, 0.1 mg/ml of BSA, and 1 mM DTT) at theindicated temperature for 10 min. Reactions were terminatedwith 5 �l of stop solution (15% Ficoll, 50 mM EDTA, 1% SDS,and 0.1% bromphenol blue). The products were analyzed by15% PAGE with 0.5� TBE. FITC-labeled DNA was visualizedby a Typhoon Trio� image analyzer (GE Healthcare).

Limited Proteolysis—TkoHef and TkoHef�ID (0.3 mg/ml ofprotein) were incubated with trypsin (Sigma) at a 1:500 enzyme:substrate (w/w ratio) in 50 mM Tris-HCl, pH 8.0, 50 mM NaCl at37 °C for 1, 3, 10, and 20 min. Reactions were stopped by theaddition of PMSF at a final concentration of 10 mM. Aliquotswere subjected to 12.5% SDS-PAGE followed by staining withCoomassie Brilliant Blue.

Circular Dichroism Measurements—The CD spectra werecollected on a J-820 spectropolarimeter (JASCO, Tokyo, Japan)at room temperature, between 190 and 250 nm. The time con-stant, scan speed, bandwidth/resolution, and sensitivity of the

spectropolarimeter were set at 1 s, 50 nm/min, 1 nm, and 100mdeg, respectively. A solution of 20 �M protein in 50 mM

sodium phosphate buffer, pH 6.8, was measured at 190 –250nm. The protein solutions in 1, 4, and 8 M urea were measuredat 210 –250 nm. The presence of urea prevented the measure-ment below 210 nm due to the high absorbance of the solution.The quantity of �-helix is evaluated from the signal at 222 nmusing Equation 1 (28).

Helicity �%� � �[�]222 � 3000�/��39000� (Eq. 1)

NMR Analysis—NMR experiments were performed on aBruker AvanceIII 600 MHz NMR spectrometer equipped witha cryogenic probe and pulsed-field gradients. To assess whetherthe IDR of the TkoHef protein (Hef-ID) exhibits properties ofIDPs, we employed a “chimeric membrane protein”-basedNMR method.3 In brief, a Pyrococcus horikoshii PH0471 mem-brane protein was engineered, and its inter-domain linkerregion (residues 72– 83) was replaced with Hef-ID. The chime-ric membrane protein PH0471-(Hef-ID) was expressed inE. coli, purified, and solubilized in dodecyl maltoside micelles.For the NMR experiments, 0.8 mg of [15N]PH0471-(Hef-ID)was dissolved in 0.30 ml of H2O/D2O (9:1), containing 25 mM

sodium phosphate buffer, pH 5.0, 100 mM NaCl, and 0.1% (w/v)dodecyl maltoside micelles. HSQC spectra (29) modified withgradient sensitivity enhancement (30) were acquired with 32transients and 256 increments at 303 K. The control spectra(PH0471 wild type (WT) and NfeD domain only) were bothrecorded with a Bruker DRX 500 MHz spectrometer at 298 K.All two-dimensional spectra were processed with nmrPipe (31)and analyzed with the program nmrDraw (31).

High Speed Atomic Force Microscopy (AFM) Imaging—A lab-oratory-built high speed AFM was used for observations of theTkoHef proteins. The instrument was previously described indetail (32, 33). AFM images were acquired in the tapping mode.Cantilevers were specially designed for high speed imaging, andprovided by Olympus. The resonant frequency was 0.8 –1.2MHz in water, and the spring constant was 0.1– 0.2 newtonm�1. The quality factor of the cantilevers in water was about 2.An amorphous carbon tip, additionally grown on the tip of acantilever by electron beam deposition, was used as the AFMprobe. For high speed AFM observations, a droplet (2 �l) of thesolution containing 1 nM TkoHef (WT or mutant proteins)was deposited on a freshly cleaved mica surface. After incuba-tion for 3 min, the samples that were not attached to the micasurface were removed by rinsing with buffer solution, and theAFM observation was then performed with the same buffersolution. The buffer solutions used for observations of the WTTkoHef and the other mutants were 300 mM NaCl, 10 mM

HEPES-NaOH, pH 7.4, and 30 mM NaCl, 10 mM HEPES-NaOH,pH 7.4, respectively. The high salt condition was important tokeep the dimerized form of the WT TkoHef on the mica sur-face. The heights and the end-to-end length analyses for deter-mining the molecular features were performed as previouslydescribed (34).

3 N. Goda, K. Shimizu, Y. Kuwahara, T. Tenno, T. Noguchi, T. Ikegami, M. Ota,and H. Hiroaki, submitted for publication.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Immunoprecipitation Experiment—A 20-�l portion ofrProtein A Sepharose Fast Flow (GE Healthcare Biosciences)was washed three times with PBS-T (10 mM sodium phosphate,pH 7.5, 150 mM NaCl, 0.1% Tween 20), mixed with 10 �l ofanti-TkoPCNA1 (35) or anti-TkoHef antiserum (5), and incu-bated at room temperature for 1 h on a rotary shaker. Eachmixture was washed twice with PBS-T, and then twice with 0.2M triethanolamine, pH 8.0. The antibody was cross-linked toProtein A with dimethyl suberimidate 2/HCl (Pierce), accord-ing to the manufacturer’s protocol. For the negative control,preimmune antiserum was used. T. kodakarensis cells (2.5 �1011 cells) were disrupted by sonication in 15 ml of lysis buffer(50 mM Tris-HCl, pH 8.0, 50 mM NaCl, 0.5 mM dithiothreitol,0.1 mM EDTA, 0.1% Triton X-100, and 10% glycerol) containingproteinase inhibitor (CompleteTM, Roche Applied Science),and the extract was obtained by centrifugation. After equilibra-tion of the antibody-conjugated rProtein A Sepharose with lysisbuffer, a 400-�l aliquot of T. kodakarensis cell extract wasadded, and the mixture was incubated for 30 min on a rotaryshaker. The precipitates were washed three times with lysisbuffer, and the immunoprecipitated proteins were eluted with40 �l of gel loading solution (50 mM Tris-HCl, pH 6.8, 1% glyc-erol, 5% �-mercaptoethanol, 0.2% bromphenol blue, 2% SDS).The protein samples were separated by SDS-12% PAGE, elec-troblotted onto a PVDF membrane, and subjected to Westernblot analysis. Proteins were visualized by an enhanced chemi-luminescence system (Millipore) and LAS-3000 mini-imageanalyzer (Fujifilm).

Surface Plasmon Resonance (SPR) Analysis—A Biacore J (GEHealthcare) system was used to study the physical interactionsbetween the proteins. All measurements were conducted at25 °C. Purified recombinant TkoPCNA1 (35), TK0155 (HAN),and His-ID were bound to CM5 sensor chips individually,according to the manufacturer’s recommendations. The con-centrations of the analyte proteins and the compositions of therunning buffer are described in the respective sections. At theend of each cycle, the bound proteins were removed by washingwith 2 M sodium chloride. The apparent equilibrium constants(KD) were determined from the association and dissociationcurves of the sensorgrams, using the BIAevaluation program(GE Healthcare).

Two-hybrid Assay—A yeast two-hybrid (Y2H) detection sys-tem (MatchmakerTM Gold Yeast Two-hybrid System, Match-maker GAL4 Two-Hybrid System 3, Clontech) was used toscreen for Hef-ID-interacting proteins. The plasmid pGBKT7,encoding the GAL4 DNA binding region, and the plasmidpGADT7, encoding the activation domains, were used to pre-pare plasmids containing the gene encoding Met490–Lys593 ofTkoHef and various genomic DNA fragments, respectively.The gene fragment encoding Met490–Lys593 was amplifiedfrom T. kodakarensis DNA using the two primers includingNdeI and NotI recognition sequences at their 5�-regions(shown in Table 1), for insertion into the NdeI-NotI sites ofpGBKT7. For the preparation of a prey library, T. kodakarensisDNA was fragmented by sonication (Microson Ultrasonic CellDisruptor, Misonix) and 0.5–1.5-kb DNA fragments were frac-tionated and eluted from an agarose gel. These DNA fragmentswere subjected to blunting with T4 DNA polymerase and E. coli

Klenow fragment, ligated with the EcoRI-NotI-BamHI adaptor,and inserted into the EcoRI site of pGADT7, after cleavage byEcoRI. The ligation solution was mixed with E. coli MegaXDH10B, and the cells were transformed by electroporation. Thesize of the prepared library was 3.1 � 106. Co-transformationsof the yeast Y2H Gold cells with pGBKT7-Hef ID and thepGADT7-library were performed according to the manufactu-rer’s protocol (Clontech Matchmaker manual). Cells were cul-tured on SD plates containing Aureobasidin A without Leu andTrp at 30 °C for 3 days for selection. The cell suspensions fromthe obtained colonies were spotted onto SD plates containingAureobasidin A with or without His. The plates were incubatedat 30 °C, and the growth of the yeast cells was monitored every24 h for 4 days. The cell growth on the plates indicated theinteractions of the two proteins produced from the two plas-mids used for the co-transformation.

RESULTS

Prediction of the IDRs in the Hef/FANCM Proteins—Thesequences of the helicase and nuclease domains are conservedin Archaea and Eukarya through evolution, but the centralregions are highly divergent in the Hef/FANCM proteins. Theamino acid sequence analysis of the archaeal Hefs searched inthe GTOP database clearly predicted that the inter-domainregion (residues 493 to 597, 105 amino acids) is entirely disor-dered (Fig. 1A). The location of the disordered region betweenthe two distinct domains was also predicted for the eukaryoticFANCM proteins by DICHOT. Their unstructured regions arequite large (about 1,243 amino acids long, including residues599 to 1,841 for human FANCM) because the eukaryoticFANCMs have long peptide chains (2,048 amino acids) (Fig.1B). These long disordered structures in the Hef/FANCM pro-teins should have physiological meanings. The sequences of thearchaeal Hef proteins in the public database were subjected tothe domain analysis, which revealed that the structural compo-sition, helicase-ID-nuclease, is conserved in Hef proteins foundin Euryarchaeota (Fig. 1C). It is evolutionally interesting thatthe other phyla of Archaea have only the nuclease domain (Cre-narchaeota) or the helicase domain (Nanoarchaeota), or sepa-rated helicase and nuclease domains (Thaumarchaeota), andthe IDR is found in the C-terminal regions of the helicasedomain (Fig. 1C).

Protein Purifications—To characterize Hef-ID, the recombi-nant IDR protein with the N-terminal His tag, His-ID, was suc-cessfully overproduced. The protein was purified to nearhomogeneity by the three sequential chromatography stepsdescribed in Fig. 2A. From a 1-liter culture (3.4 g cells), 1.49 mgof homogeneous protein was obtained. In addition, we purifiedtwo Hef mutant proteins that are deficient in dimer formation,TkoHef-m (based on knowledge from our structural analyses ofP. furiosus Hef (36)) and TkoHef�ID, with a deletion of theIDR(493–588), as well as the WT Hef protein. The purifiedproteins are shown in Fig. 2, B and C.

Limited Proteolysis—Accessibility and flexibility of the sub-strate affect on the cleavage efficiency of the proteases, and,disordered regions are generally easy targets for proteolyticattack. Therefore, using purified TkoHef and TkoHef�ID asdescribed above, trypsin reaction was performed with a limited




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

condition. As shown in Fig. 3A, the WT TkoHef was cleaved bytrypsin to two major parts with increasing reaction time. On theother hand, TkoHef�ID showed a highly resistant feature to thetrypsin reaction (Fig. 3B). The cleavage products of TkoHefcorrespond to the sizes of the helicase and nuclease domains,respectively (Fig. 3C). Several bands were observed around eachcleaved product (Fig. 3A), suggesting that TkoHef was cleavedat several different sites in the ID region, which contains manyLys and Arg residues. These results indicate that TkoHef hastrypsin-susceptible sites at the predicted ID region.

Structural Analyses of the Hef-ID—The CD spectrum of thepurified His-ID protein was recorded in the wavelength rangeof 190 –250 nm (Fig. 4A). The spectrum in a phosphate buffersolution was characterized by a shoulder at 222 nm and anintense negative minimum at 202 nm. These features revealthat His-ID is mostly disordered, but contains some helicalstructure (7%), which was calculated as described earlier (28),followed by subtraction of the value in the case of the spectrumwith 8 M urea. For further structural analysis of the disorderedproperty of Hef-ID by NMR, we used our chimeric membraneprotein-based systematic NMR assessment system for IDP. Inthis system, the target protein of interest (Hef-ID in this study)was expressed as a fusion with an archaeal membrane protein,PH0471, which provides a well known HSQC spectrum, as

FIGURE 1. IDR in the TkoHef and FANCM proteins. The classifications of TkoHef in the GTOP database (ID code BAD85210.1) (A) and human FANCM byDICHOT prediction (ID code FANCM_HUMAN) (B) are presented. The structural regions with similarity to known three-dimensional structures are shown withPDB codes under the bars, and unstructured regions are colored black. The numbers above the black bars indicate the amino acid residues of unstructuredregions. The domains are shown by the arrows. C, domain analyses of archaeal Hef homologs from each subdomain are shown. The helicase and nucleasedomains are shown in gray and white, respectively. IDRs are black. The classification of unstructured regions is according to the DISOPRED lines in the GTOPdatabase. GenBank accession numbers used in GTOP are as follows. T. kodakarensis, BAD85210.1; P. furiosus, AAL82139.1; P. abysii, CAB49203.1; Methanosarcinamazei, AAM31083.1; Archaeoglobus fulgidus, AAB89786.1; Methanococcus jannaschii, AAB99518.1; Halobacterium salinarum, CAP14762.1; Methanothermobac-ter thermautotrophicus, AAB85892.1; Thermoplasma acidophilum, Ta1501.1; Sulfolobus solfataricus, eiF4A; Candidatus K. cryptofilum, ACB08030.1; Nanoar-chaeum equitanss, AAR39218.1; Cenarchaeum symbiosum, ABK77806.1, ABK76661.1; Nitrosopumlis maritimus, ABX128401.1, ABX12333.1.

FIGURE 2. Purification of TkoHef and related proteins in this study. A,purified His-ID (1 �g) was fractionated by 12.5% SDS-PAGE. B, TkoHef andTkoHef�ID (1 �g each) were fractionated by 7% SDS-PAGE on the same gel. C,TkoHef-m (1 �g) was fractionated by 10% SDS-PAGE. The proteins werestained by Coomassie Brilliant Blue. Protein markers (lane 1) were obtainedfrom New England Biolabs Inc.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

described under “Experimental Procedures.” In the 1H-15NHSQC spectrum of the PH0471-Hef-ID fusion protein, 73 sig-nals of the 179 observed signals (Fig. 4B) were exactly the sameas those from the PH0417 protein by itself (Fig. 4D), mostlycorresponding to the signals from the isolated C-terminal NfeDdomain (Fig. 4C). This result suggested that the Hef-ID part washighly flexible in solution. In addition, 89 additionally observedsignals originating from NH groups of the Hef-ID (96%) showedthe limited chemical shift dispersion within 8.6 and 7.8 ppm(Fig. 4B). These results clearly supported our prediction thatthe IDR of TkoHef in the fusion protein is predominantlyunstructured and flexible.

The IDR Contributes to the Flexible Motion of the Two Dis-tinct Domains of TkoHef—AFM visualizes individual proteinmolecules directly as a frame shot of time-averaged features. Incombination with a high speed recording technique, images ofthe structural changes of a single molecule can be captured.This technique was used to observe the TkoHef protein todetermine whether the IDR is actually disordered in theTkoHef protein. In the case of WT TkoHef, two large globuleswere connected to a slightly smaller globule by thin flexiblelinkers, and the globules, especially the smaller one, moved rap-idly, as shown in Fig. 5A (see supplemental Movie AFM-S1).The images from TkoHef-m revealed that only one large glob-

FIGURE 3. Limited proteolysis of TkoHef and TkoHef�ID with trypsin.TkoHef (A) and TkoHef�ID (B), 0.3 mg/ml protein, were incubated withtrypsin at 1:500 enzyme:substrate (w/w ratio) in 50 mM Tris-HCl, pH 8.0, 50mM NaCl at 37 °C. Lane 1, molecular weight markers; lanes 2– 6, samplesafter 0, 1, 3, 10, and 20 min. C, helicase and nuclease domains are shown bygray and white bars, respectively. IDR is shown by a black bar. The calcu-lated molecular weights corresponding to each part are shown on thelines with arrowheads.

FIGURE 4. Conformational properties of the IDR, analyzed by far-UV CD and NMR. A, far-UV CD spectra of His-ID protein. Twenty �M protein was measuredin 50 mM sodium phosphate, pH 6.8. Spectra of the protein in 1, 4, and 8 M urea at 210 –250 nm were superimposed. B, 1H-15N HSQC spectra of Hef-ID fused withthe membrane protein, PH0471, from P. horikoshii. C, the HSQC spectra of the only NfeD domain of PH0471. D, the HSQC spectra of PH0471.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

ule was connected to a small globule by a thin flexible linker,and the small globule moved rapidly, as shown in Fig. 5B (seesupplemental Movie AFM-S2). These results demonstratedthat the TkoHef protein forms a dimer through the small glob-ule. According to the crystal structure of P. furiosus Hef (36, 37),the helicase domain is larger than the nuclease domain. Thus,the large and small globules represent the helicase and nucleasedomains of TkoHef, respectively. These assignments are alsosupported by our previous report showing that P. furiosus Hefformed a dimer at the nuclease domain (2). Note that the char-acteristic “c”-like shape of the crystal structure of the helicasedomain (37) was sometimes observed in the large globule (forexample, at 0.40 and 4.22 s in Fig. 5B), providing further evi-dence that the large globule is the helicase domain. On theother hand, in the case of TkoHef�ID, the large and small glob-ules were connected to each other, and no flexible linker wasobserved, as shown in Fig. 5C (see supplemental Movie AFM-S3). The heights of the large and small globules were 3.6 0.5(n � 2,717) and 3.0 0.6 nm (n � 2,717) for TkoHef-m (Fig. 6,D and E), and 3.6 0.4 (n � 1,317) and 2.9 0.4 nm (n � 1,317)for TkoHef�ID (Fig. 6, G and H), respectively, which are verysimilar to each other. The only visible difference between theAFM images of the two constructs was the existence of theflexible linker. Thus, it can be concluded that the flexible linkerseen in TkoHef-m is the IDR of TkoHef. On the other hand, theheight of the large globule was 4.4 0.6 nm (n � 4,540) for

TkoHef (Fig. 6A), which is higher than those of TkoHef-m andTkoHef�ID. This height variation may be caused by the differ-ent observation conditions. The observations of TkoHef wereperformed under conditions with high salt buffer, which oftenweakens the interactions between proteins and the mica sur-face, and thus slightly increases the height of proteins due toenhanced structural flexibility. The height of the small globuleof TkoHef is 3.6 0.6 nm (n � 2,270, Fig. 6B), which is alsohigher than those of TkoHef-m and TkoHef�ID. This may becaused by the combined effects of the different observation buf-fers and the dimerization of this domain. Furthermore, themean end-to-end lengths of the flexible linker for the WT andTkoHef-m proteins were 14 5 (n � 4,540) and 14 6 nm (n �2,717), respectively (Fig. 6, C and F). Because these values aresimilar to each other, the flexible linker detected in the WTmust also be the ID region. The maximum end-to-end length,which may be very close to the contour length, of the ID regionfor the WT and monomer mutant were 32 (among 4,540frames) and 35 nm (among 2,717 frames), respectively (Fig. 6, Cand F). Assuming that the average distance between the nearestneighbor amino acids is 0.34 nm, and that the ID region ofTkoHef is composed of 104 amino acids, these values seem tobe quite reasonable (0.34 nm � 104 � 35.4 nm). The parame-ters obtained here may not reach the most probable values,because only three of the typical molecules for each constructusing 1,317–2,717 frames were analyzed here. However, we

FIGURE 5. Typical high speed AFM images and schematic illustrations showing the molecular features of three TkoHef proteins. A, WT (seesupplemental Movie AFM-S1). B, TkoHef-m (dimer formation-defective mutant) (see supplemental Movie AFM-S2). C, TkoHef�ID (IDR deletion mutant)(see supplemental Movie AFM-S3). The time from the beginning of the movie is indicated at the lower left of each image. Scale bar, 30 nm; Z-scale, 5.0nm for A and 3.5 nm for B and C. Light blue, pink, and green arrowheads represent helicase domain, nuclease domain, and IDR of the TkoHef protein,respectively. The rightmost schematic illustrations use the same color codes as those used for the arrowheads in the AFM images. The image analysisresults are presented in Fig. 6.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

believe that these parameters are convincing for our conclu-sions because they represent well the molecular features of eachconstruct. Taken together, these AFM observations directlydemonstrated the molecular features of the three TkoHef pro-teins, and suggested that the predicted IDR actually existsbetween the helicase and nuclease domains. The IDR movedrapidly and lacked a distinct structure, which may facilitatestructural conversions for the different functions of TkoHef invarious cellular phenomena.

Interaction between TkoPCNA1 and TkoHef—We predictedthat the IDR of TkoHef functions in interactions with variouspartner molecules with different binding modes, by coupledbinding and folding, and thus we searched for the IDR-bindingproteins. PCNA is well known as a scaffold molecule that inter-acts with numerous proteins involved in DNA replication andrepair. Therefore, an immunoprecipitation experiment wasperformed using a T. kodakarensis cell extract, the antiseraagainst TkoPCNA1, one of the two PCNAs in this organism,and the antisera against TkoHef. As shown in Fig. 7A, theTkoPCNA1 band was detected in the fraction precipitated withthe anti-TkoHef antibody. Conversely, TkoHef was co-precipi-tated with TkoPCNA1 by the anti-TkoPCNA1 antibody (Fig.7B). A faint band of TkoPCNA1 was detected in the mock lane(precipitated with the preimmune antiserum). However, quan-tifications of each band in this experiment (repeated five times)showed significant differences of the band intensities betweenthe co-precipitation lanes and mock lanes. Although it wouldbe possible that these two proteins were co-precipitatedthrough DNA, but not by direct interaction, these results sup-

ported that a complex containing both TkoPCNA1 and TkeHefexists, at least in T. kodakarensis cells. The cell extract wasprepared by sonication with a condition that most of the DNAstrands were digested to very short fragments, and therefore,

FIGURE 6. Results of the high speed AFM image analysis. A, D, and G, height distributions of the large globules of TkoHef, TkoHef-m, and TkoHef�ID,respectively. B, E, and H, height distributions of the small globules of TkoHef, TkoHef-m, and TkoHef�ID, respectively. C, F, and I, distributions of end-to-endlengths between the large and small globules of TkoHef, TkoHef-m, and TkoHef�ID, respectively. These histograms were obtained from three typical moleculesfor each construct. N represents number of frames used in each histogram. Because TkoHef has two large globules, the number of frames used for A and C aretwice the number of frames analyzed.

FIGURE 7. Interaction of TkoHef with TkoPCNA1. A, T. kodakarensis cellextracts were used for the pulldown experiment with anti-TkoHef antiserum,followed by Western analysis. The band intensities of PCNA1 were quantifiedby MultiGauge software (Fuji film). B, T. kodakarensis cell extracts were pulleddown with anti-TkoPCNA1 antiserum, followed by Western analysis. Theband intensities of Hef were quantified by MultiGauge software (Fuji film).The mean S.E. for the relative intensities detected in mock lanes wereobtained from five repeats and shown under the gel images. C, SPR analyseswere performed with TkoPCNA1 as a ligand. Various protein concentrationsof TkoHef (indicated on the sensorgram) were loaded on the TkoPCNA1-bound chip for 120 s in a running buffer containing 10 mM HEPES, pH 7.4, and350 mM NaCl. Each sensorgram showed the actual binding response obtainedby subtraction of the background response.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

we thought the co-precipitation of TkoPCNA1 and TkeHefprobably resulted from their direct interaction. To confirm thedirect interaction, we performed SPR experiments with thepurified TkoHef and TkoPCNA1 proteins. The TkoPCNA1protein was covalently immobilized on the Biacore CM5 sensorchip, and different concentrations of the TkoHef protein weresubsequently injected. The SPR sensorgram clearly showed thephysical interaction between the TkoHef and TkoPCNA1 pro-teins (Fig. 7C). The calculated KD value for the interaction was5.3 � 10�7 M, which is comparable with the values for otherPCNA-binding proteins in P. furiosus (1.1 � 10�7 M for DNAligase, 9.9 � 10�8 M for DNA polymerase BI, and 2.2 � 10�7 M

for uracil DNA glycosylase), as determined by SPR analyses inour laboratory (38 – 40).

TkoPCNA1 Interacts with the IDR of TkoHef—We searchedfor the PIP-box motif sequence, which is widely conserved inPCNA-interacting proteins (41), in the sequence of TkoHef.However, no typical motif sequence was found in either the Nor C terminus of the peptide chain, where the PIP motif usuallyexists in many PCNA-binding proteins. Therefore, we pre-dicted that the PCNA-binding site of TkoHef is located in theIDR, because PCNA-binding sites also exist in structurally flex-ible areas, such as looped-out regions, when present in theinternal region of a peptide chain (38, 40). We subjectedTkoHef�ID to the same SPR analysis. As shown in Fig. 8A, noresponse was observed in the sensorgram for TkoHef�ID. Fur-thermore, the purified His-ID protein showed a clearly positivesensorgram, although the calculated KD value, 9.1 � 10�6 M,was 1 order of magnitude higher than that of the WT TkoHef(Fig. 8B). All of the data supported the proposal that TkoHefuses its IDR to bind to TkoPCNA1.

Exhaustive Search for Proteins Interacting with the IDR ofTkoHef—One of the important functions of IDR is to interactwith plural partners by converting its conformation. Therefore,we tried an exhaustive screening of the candidate proteins thatinteract with TkoHef-ID by the Y2H experiment. The geneencoding TkoHef-ID was cloned into the bait plasmidpGBKT7, as a fusion with the gene encoding the GAL4 DNAbinding domain. A genomic DNA library was constructed usingT. kodakarensis DNA and the prey plasmid pGADT7, as fusionswith the genes encoding the GAL4 activation domain, asdescribed under “Experimental Procedures.” Fourteen clonesgrew on the selection plates. The prey plasmids were then pre-pared from these positive clones, and the nucleotide sequences

of the inserted fragments were determined. A public databasesearch for the genes providing positive signals revealed eachNCBI locus tag (TK number), and 10 different genes were iden-tified among the 14 clones. In the case of four genes (TK0356,TK0467, TK1577, and TK2303) among the 14 clones, twoclones each with different regions were included. The positiveclones were cultivated individually and a portion for each liquidculture was spotted on the selection plate again (Fig. 9). Four ofthe 14 clones grew on the selection plates with the empty baitplasmid, and these self-activating clones, one of TK0356,TK0669, and both of TK1577, were excluded from the candi-dates. Most of the gene products are annotated as hypotheticalproteins, as shown in Table 2, and the functions of these pro-teins should be analyzed individually. One interesting candi-date, TK0155, was a bacterial RecJ-like protein. RecJ is a 5�–3�exonuclease involved in several DNA repair pathways. There-fore, TK0155 may be a participant in the repair process, afterHef forms a repairsome complex at the stalled replication fork.The gene encoding TK0155 was cloned, and the gene productwas produced as a full-length protein in E. coli cells (Fig. 10A).The purified TK0155 protein was confirmed to have the exo-nuclease activity, which is optimal at higher temperatures (Fig.10B). An SPR analysis was performed, using the His-ID-immo-bilized sensor chip and the purified TK0155 protein, and bind-ing of the two proteins was quantitatively confirmed, with anapparent KD value of 3.6 � 10�6 M (Fig. 11A). Based on these

FIGURE 9. Screening of the Hef-ID-interacting proteins. Interactionsbetween the TkoHef-ID and T. kodakarensis proteins were analyzed by a yeasttwo-hybrid assay. Clones providing positive signals were selected, and theprey plasmids were prepared. The proteins deduced from the DNA sequencesinserted in the prey plasmids are shown as NCBI locus tag (TK #). The transfor-mants that grew on the selection agar plates are shown. Cell suspensions (3 �lof 1.2 � 107 cell/ml) of each strain were spotted onto S.D. plates containingAureovasidin A (120 ng/ml) without Leu and Trp (middle), and Leu, Trp, andHis (bottom) for two different selection strengths. Transformants on the non-selection plate are shown at the top. Minus indicates the transformants withthe bait or prey plasmid without insert DNA. The agar plates were incubatedat 30 °C for 3 days, and growing cells were visualized.

TABLE 2Candidate proteins for Hef-ID binding

Locus tag Annotation

TK0155 RecJ-like exonucleaseTK0356 Hypothetical membrane proteinTK0467 Hypothetical protein TK0467TK0522 Carbohydrate esterase family 1 proteinTK1781 Diaminopimelate aminotransferaseTK1997 Hydrogenase maturation protein HypFTK2021 ParA/MinD family ATPaseTK2303 Chaperonin � subunit

FIGURE 8. SPR analyses of TkoHef proteins with the TkoPCNA1-boundchip. WT TkoHef and TkoHef�ID (2 �M each) (A), and His-ID (concentrationsare indicated on the sensorgram) (B) were loaded on the TkoPCNA1-boundchip for 120 s in a running buffer containing 10 mM HEPES, pH 7.4, and 350 mM

NaCl.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

experiments, we propose naming the RecJ-like protein as HAN,for Hef-associated nuclease.

PCNA and HAN (a RecJ-like Protein) Do Not Bind Hef-IDSimultaneously—The advantage of intrinsically disorderedproteins is that they adopt various forms, depending on thebinding partner. We expected that Hef-ID would fold in differ-ent manners for binding to PCNA1 and HAN. To investigate

whether the IDR binds to PCNA1 and HAN simultaneously, weperformed an SPR analysis. His-ID was loaded onto a HAN-immobilized sensor chip. As shown in Fig. 11B, His-ID boundthe immobilized HAN in the opposite combination of ligandand analyte as the case of Fig. 11A. Addition of PCNA1decreased the responses in the concentration-dependent man-ner (Fig. 11B). Furthermore, in the case of loading of His-IDonto a PCNA1-immobilized sensor chip, addition of HANdecreased the responses as well (Fig. 11C). The results of theseexperiments support the idea that different ligands could bindto the same region in the IDR.

DISCUSSION

Presence of IDR in the Hef/FANCM Proteins—Structural andfunctional analyses of human FANCM and its related proteinshave become extremely popular, since FANCM was found to bethe ortholog of the archaeal Hef protein. We focused here onthe region between the N-terminal DEAH helicase-like domainand the C-terminal XPF/Mus81 nuclease-like domain of theHef/FANCM proteins, because of their disordered structures.In general, IDR have a characteristically skewed amino acidcomposition, with an abundance of hydrophilic and chargedresidues, and few hydrophobic residues (42, 43). The IDR arealso characterized by higher sequence divergence, as comparedwith the structural domains. The structural domains are wellconserved through evolution, and can be detected by homologysearches even across different domains of life (44, 45). By con-trast, the IDR generally mutate more rapidly than the structuralregions, presumably because the IDR are not structurally con-strained against frequent insertions/deletions and amino acidsubstitutions. Indeed, the cases of the Hef/FANCM proteinsrevealed that the inter-domain regions are highly divergent, incontrast to the two distinct domains. All of our structural anal-yses presented in this study showed that the inter-domainregion in TkoHef is disordered in solution. Furthermore, weidentified two proteins, PCNA1 and HAN, as binding partnersof TkoHef at its IDR. We only present here that PCNA1 andHAN do not bind Hef-ID simultaneously from the experimen-tal data showing failure to detect a ternary complex of Hef-ID�PCNA1�HAN, and the binding mode for each partnerremains to be determined. We predict that a part or the entireregion (104 amino acids long) of Hef-ID will be flexibly foldeddepending on its binding partners. The adoption of differentstructures will express the binding specificity to each partner.Although the sequences are divergent among species, their dis-ordered structures would be conserved in the Hef/FANCMproteins.

FANCM-interacting Proteins—Human FANCM interactswith many proteins to form an FA core complex (FANCA,FANCB, FANCC, FANCE, FANCF, FANCG, FANCL,FAAP100, FAAP24, and HES1) on the stalled replication fork.The inter-domain region of human FANCM (1,243 aminoacids) is 10 times longer than that in T. kodakarensis. The entiredisorder of this inter-domain region can be predicted from itsamino acid sequence, and it suggests chances to interact withmuch more proteins to process a more complex repair systemin Eukarya as compared with Archaea. In the vertebratehomologs of FANCM, three highly conserved sequence motifs

FIGURE 10. Purification of TK0155 (HAN) and nuclease assay. A, purifiedTK0155 (1 �g) was fractionated by 10% SDS-PAGE and stained by CoomassieBrilliant Blue. B, temperature-dependent exonuclease activity of the TK0155protein. A single-stranded DNA labeled with FITC at the 3� terminus was incu-bated with TK0155 (HAN) at 37, 55, and 70 °C. The products were fractionatedby 15% PAGE in 0.5� TBE.

FIGURE 11. PCNA and HAN do not bind Hef-ID simultaneously. A, purifiedHAN, at the concentrations indicated beside the sensorgram, was loaded onthe His-ID-bound chip for 120 s at flow rate 30 �l/min in the running buffer, 10mM HEPES, pH 7.4, 150 mM NaCl, and 10% Tween 20. B, the His-ID protein (0.5�M) was loaded on the HAN-bound chip for 60 s at flow rate 15 �l/min in thesame running buffer as A. Different concentrations (indicated with a � (�M)on the sensorgram) of PCNA1 were mixed with 0.5 �M His-ID before loading.The dotted line indicates the injection of 4 �M PCNA1 as a negative control. C,the His-ID protein (0.5 �M) was loaded on the PCNA1-bound chip for 60 s atflow rate 15 �l/min in the same running buffer as A. Different concentrations(indicated with a � (�M) on the sensorgram) of HAN were mixed with 0.5 �M

His-ID before loading. The dotted line indicates the injection of 2 �M HAN as anegative control.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

were identified in the inter-domain region and designated asMM1, MM2, and MM3 (FANCM motifs 1–3), and these motifswere predicted to be important for protein-protein interactions(46). Biochemical analyses revealed that MM1 interacts withthe FA core complex by binding to FANCF, whereas MM2interacts with RM1 and topoisomerase IIIa, components of theBloom’s syndrome complex. Furthermore, MM1 and MM2were independently required to activate the FA and Bloom’ssyndrome repair pathways. Therefore, the proper function ofFANCM is responsible for preventing both Fanconi anemia andBloom’s syndrome. In addition, two histone-fold containingproteins, MHF1 and MHF2, which were recently identified asFANCM-associated factors, form the stable complex MHF1-MHF2 and stimulate the DNA-binding activity of FANCM (47,48). The stable association with FANCM and the DNA-bindingactivity of MHF activate the FA repair pathway. The bindingsite of the MHF1�MHF2 complex, located between the DHAFhelicase domain and MM1 (the region composed of aminoacids 661– 800) in FANCM is also predicted to be disordered(48). The crystal structure of an MHF1-MHF2-FANCM frag-ment was published quite recently (49). These findings indi-cated that the inter-domain region in FAMCM is importantfor the interactions with several proteins in processing to thedownstream reactions in the repair pathways, and the disor-dered structure of this region is suitable for accomplishingmultiple functions of FANCM with its various interactivepartners.

PCNA Interacts with Hef—We predicted that PCNA mayinteract with Hef, because of its function at the replication fork.Indeed, one of the two PCNA proteins in T. kodakarensis boundto Hef at IDR. The bound PCNA was PCNA1, which is essentialfor the viability of T. kodakarensis and is probably a regularmember of the replisome (35). PCNA1 in the progressing repli-some can stimulate recruitment of Hef when the replicationfork stalls. A recent report describing the in vivo network anal-yses of the Pyrococcus abyssi cells showed that the Hef homologis included in a His-tagged PCNA-bound fraction (50). Thisresult also supports the interaction between PCNA and Hef inArchaea. From our experience, interaction with the PCNAprotomer is not easy to detect by Y2H methods for an unknownreason, and even a homologous interaction of PCNA-PCNA(for its ring formation) cannot be detected. Therefore, the factthat the positive clones did not include PCNA from the Y2Hexperiment in this study is understandable. In T. kodakarensis,PCNA1 is essential and the major clamp molecule in DNA rep-lication. On the other hand, the gene for PCNA2 can beknocked out and the physiological function of PCNA2 has notbeen elucidated. Furthermore, many more molecules ofPCNA1 exist as compared with PCNA2 (35). We analyzedonly PCNA1 in this study, and will see the relationshipsbetween PCNA2 and Hef in our further works. In Eukarya,post-translational modifications of PCNA, including phos-phorylation, acetylation, ubiquitination, and SUMOylation(51), regulate its binding to interacting proteins. However,neither ubiquitination nor SUMOylation have been reportedin Archaea, and it is not clear if any post-translational mod-ifications of PCNA are involved in the Hef-related repairprocess.

Identification of HAN, a RecJ-like Nuclease, as a Hef-interact-ing Protein—We obtained eight proteins as TkoHef-interactingproteins by the Y2H screening. Among these, HAN, a RecJ-likeprotein, is interesting, because it displayed an exonucleaseactivity. The concrete role of exonuclease activity of this pro-tein remains to be investigated, but it should act downstream ofincision at the stalled fork to start homologous recombinationor remove lesions by a nucleotide excision repair-like process. Itis interesting that two RecJ-like proteins are present in T. koda-karensis and other archaeal organisms. One of them interactswith Hef (in this study) and the other interacts with GINS, anessential factor for DNA replication initiation and elongation inEukarya and Archaea, which may be a component of the repli-cative helicase complex. The latter RecJ-like protein wasreported as the GINS-associated nuclease (GAN), and may beinvolved in lagging strand processing (52). Furthermore, char-acterization of the GAN ortholog from P. furiosus revealed the3�–5� exoribonuclease activity, which may function in proof-reading for primer synthesis (53). The work sharing the twoRecJ-like proteins, GAN and HAN, in Archaea is an interestingsubject for future analyses.

Acknowledgments—We thank Drs. Kosuke Morikawa and MamoruSato for valuable discussions and encouragement. We also thank Drs.Daisuke Kohda and Yoshito Abe for the CD spectra analysis.

REFERENCES1. Komori, K., Fujikane, R., Shinagawa, H., and Ishino, Y. (2002) Novel en-

donuclease in Archaea cleaving DNA with various branched structure.Genes Genet. Syst. 77, 227–241

2. Komori, K., Hidaka, M., Horiuchi, T., Fujikane, R., Shinagawa, H., andIshino, Y. (2004) Cooperation of the N-terminal helicase and C-terminalendonuclease activities of archaeal Hef protein in processing stalled rep-lication forks. J. Biol. Chem. 279, 53175–53185

3. Nishino, T., Ishino, Y., and Morikawa, K. (2006) Structure-specific DNAnucleases: structural basis for 3D-scissors. Curr. Opin. Struct. Biol. 16,60 – 67

4. Sato, T., Fukui, T., Atomi, H., and Imanaka, T. (2005) Improved and ver-satile transformation system allowing multiple genetic manipulations ofthe hyperthermophilic archaeon Thermococcus kodakaraensis. Appl. En-viron. Microbiol. 71, 3889 –3899

5. Fujikane, R., Ishino, S., Ishino, Y., and Forterre, P. (2010) Genetic analysisof DNA repair in the hyperthermophilic archaeon, Thermococcus kodak-araensis. Genes Genet. Syst. 85, 243–257

6. Lestini, R., Duan, Z., and Allers, T. (2010) The archaeal Xpf/Mus81/FANCM homolog Hef and the Holliday junction resolvase Hjc definealternative pathways that are essential for cell viability in Haloferax volca-nii. DNA Repair 9, 994 –1002

7. Lestini, R., Laptenok, S. P., Kuhn, J., Hink, M. A., Schanne-Klein, M. C.,Liebl, U., and Myllykallio, H. (2013) Intracellular dynamics of archaealFANCM homologue Hef in response to halted DNA replication. NucleicAcids Res. 41, 10358 –10370

8. Roberts, J. A., Bell, S. D., and White, M. F. (2003) An archaeal XPF repairendonuclease dependent on a heterotrimeric PCNA. Mol. Microbiol. 48,361–371

9. Newman, M., Murray-Rust, J., Lally, J., Rudolf, J., Fadden, A., Knowles,P. P., White, M. F., and McDonald, N. Q. (2005) Structure of an XPFendonuclease with and without DNA suggests a model for substrate rec-ognition. EMBO J. 24, 895–905

10. Jones, R. M., and Petermann, E. (2012) Replication fork dynamics and theDNA damage response. Biochem. J. 443, 13–26

11. Errico, A., and Costanzo, V. (2012) Mechanisms of replication fork pro-




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

tection: a safeguard for genome stability. Crit. Rev. Biochem. Mol. Biol. 47,222–235

12. Sengerova, B., Wang, A. T., and McHugh, P. J. (2011) Orchestrating thenucleases involved in DNA interstrand cross-link (ICL) repair. Cell Cycle10, 3999 – 4008

13. Franchitto, A., and Pichierri, P. (2011) Understanding the molecular basisof common fragile sites instability: role of the proteins involved in therecovery of stalled replication forks. Cell Cycle 10, 4039 – 4046

14. Kottemann, M. C., and Smogorzewska, A. (2013) Fanconi anaemia and therepair of Watson and Crick DNA crosslinks. Nature 493, 356 –363

15. Kim, H., and D’Andrea, A. D. (2012) Regulation of DNA cross-link repairby the Fanconi anemia/BRCA pathway. Genes Dev. 26, 1393–1408

16. Constantinou, A. (2012) Rescue of replication failure by Fanconi anaemiaproteins. Chromosoma 121, 21–36

17. Crossan, G. P., and Patel, K. J. (2012) The Fanconi anaemia pathwayorchestrates incisions at sites of crosslinked DNA. J. Pathol. 226,326 –337

18. Deans, A. J., and West, S. C. (2011) DNA interstrand crosslink repair andcancer. Nat. Rev. Cancer 11, 467– 480

19. Mosedale, G., Niedzwiedz, W., Alpi, A., Perrina, F., Pereira-Leal, J. B.,Johnson, M., Langevin, F., Pace, P., and Patel, K. J. (2005) The vertebrateHef ortholog is a component of the Fanconi anemia tumor-suppressorpathway. Nat. Struct. Mol. Biol. 12, 763–771

20. Meetei, A. R., Medhurst, A. L., Ling, C., Xue, Y., Singh, T. R., Bier, P.,Steltenpool, J., Stone, S., Dokal, I., Mathew, C. G., Hoatlin, M., Joenje, H.,de Winter, J. P., and Wang, W. (2005) A human ortholog of archaeal DNArepair protein Hef is defective in Fanconi anemia complementation groupM. Nat. Genet. 37, 958 –963

21. Tompa, P. (2012) Intrinsically disordered proteins: a 10-year recap. TrendsBiochem. Sci. 37, 509 –516

22. Babu, M. M., van der Lee, R., de Groot, N. S., and Gsponer, J. (2011)Intrinsically disordered proteins: regulation and disease. Curr. Opin.Struct. Biol. 21, 432– 440

23. Uversky, V. N., Oldfield, C. J., Midic, U., Xie, H., Xue, B., Vucetic, S.,Iakoucheva, L. M., Obradovic, Z., and Dunker, A. K. (2009) Unfoldomicsof human diseases: linking protein intrinsic disorder with diseases. BMCGenomics 10, S7

24. Rezaei-Ghaleh, N., Blackledge, M., and Zweckstetter, M. (2012) Intrinsi-cally disordered proteins: from sequence and conformational propertiestoward drug discovery. Chembiochem 13, 930 –950

25. Fukuchi, S., Homma, K., Sakamoto, S., Sugawara, H., Tateno, Y., Gojobori,T., and Nishikawa, K. (2009) The GTOP database in 2009: updated con-tent and novel features to expand and deepen insights into protein struc-tures and functions. Nucleic Acids Res. 37, D333-D337

26. Fukuchi, S., Homma, K., Minezaki, Y., Gojobori, T., and Nishikawa, K.(2009) Development of an accurate classification system of proteins intostructured and unstructured regions that uncovers novel structural do-mains: its application to human transcription factors. BMC Struct. Biol. 9,26

27. Gill, S. C., and von Hippel, P. H. (1989) Calculation of protein extinc-tion coefficients from amino acid sequence data. Anal. Biochem. 182,319 –326

28. Morrisett, J. D., David, J. S., Pownall, H. J., Gotto, A. M. Jr. (1973) Interac-tion of an apolipoprotein (apoLP-alanine) with phosphatidylcholine. Bio-chemistry 12, 1290 –1299

29. Neri, D., Szyperski, T., Otting, G., Senn, H., and Wuthrich, K. (1989)Stereospecific nuclear magnetic resonance assignments of the methylgroups of valine and leucine in the DNA-binding domain of the 434 re-pressor by biosynthetically directed fractional 13C labeling. Biochemistry28, 7510 –7516

30. Forman-Kay, J. D., Clore, G. M., and Gronenborn, A. M. (1992) Relation-ship between electrostatics and redox function in human thioredoxin:characterization of pH titration shifts using two-dimensional homo- andheteronuclear NMR. Biochemistry 31, 3442–3452

31. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G., Pfeifer, J., and Bax, A.(1995) NMRPipe: a multidimensional spectral processing system based onUNIX pipes. J. Biomol. NMR 6, 277–293

32. Ando, T., Kodera, N., Takai, E., Maruyama, D., Saito, K., and Toda, A.

(2001) A high-speed atomic force microscope for studying biological mac-romolecules. Proc. Natl. Acad. Sci. U.S.A. 98, 12468 –12472

33. Ando, T., Uchihashi, T., and Fukuma, T. (2008) High-speed atomic forcemicroscopy for nano-visualization of dynamic biomolecular processes.Prog. Surf. Sci. 83, 337–347

34. Hashimoto, M., Kodera, N., Tsunaka, Y., Oda, M., Tanimoto, M., Ando,T., Morikawa, K., and Tate, S. (2013) Phosphorylation-coupled intramo-lecular dynamics of unstructured regions in chromatin remodeler FACT.Biophys. J. 104, 2222–2234

35. Kuba, Y., Ishino, S., Yamagami, T., Tokuhara, M., Kanai, T., Fujikane,R., Daiyasu, H., Atomi, H., and Ishino, Y. (2012) Comparative analysesof the two proliferating cell nuclear antigens from the hyperthermo-philic archaeon, Thermococcus kodakarensis. Genes Cells 17, 923–937

36. Nishino, T., Komori, K., Ishino, Y., and Morikawa, K. (2003) X-ray andbiochemical anatomy of an archaeal XPF/Rad1/Mus81 family nuclease:similarity between its endonuclease domain and restriction enzymes.Structure 11, 445– 457

37. Nishino, T., Komori, K., Tsuchiya, D., Ishino, Y., and Morikawa, K. (2005)Crystal structure and functional implications of Pyrococcus furiosus Hefhelicase domain involved in branched DNA processing. Structure 13,143–153

38. Kiyonari, S., Takayama, K., Nishida, H., and Ishino, Y. (2006) Identifica-tion of a novel binding motif in Pyrococcus furiosus DNA ligase for thefunctional interaction with proliferating cell nuclear antigen. J. Biol. Chem.281, 28023–28032

39. Tori, K., Kimizu, M., Ishino, S., and Ishino, Y. (2007) DNA polymerases BIand D from the hyperthermophilic archaeon Pyrococcus furiosus bothbind to proliferating cell nuclear antigen with their C-terminal PIP-boxmotifs. J. Bacteriol. 189, 5652–5657

40. Kiyonari, S., Uchimura, M., Shirai, T., and Ishino, Y. (2008) Physical andfunctional interactions between uracil-DNA glycosylase and proliferatingcell nuclear antigen from the euryarchaeon Pyrococcus furiosus. J. Biol.Chem. 283, 24185–24193

41. Warbrick, E. (2000) The puzzle of PCNA’s many partners. Bioessays 22,997–1006

42. Dunker, A. K., Lawson, J. D., Brown, C. J., Williams, R. M., Romero, P., Oh,J. S., Oldfield, C. J., Campen, A. M., Ratliff, C. M., Hipps, K. W., Ausio, J.,Nissen, M. S., Reeves, R., Kang, C., Kissinger, C. R., Bailey, R. W., Griswold,M. D., Chiu, W., Garner, E. C., and Obradovic, Z. (2001) Intrinsicallydisordered protein. J. Mol. Graph. Model 19, 26 –59

43. Weathers, E. A., Paulaitis, M. E., Woolf, T. B., and Hoh, J. H. (2004) Re-duced amino acid alphabet is sufficient to accurately recognize intrinsi-cally disordered protein. FEBS Lett. 576, 348 –352

44. Vogel, C., Berzuini, C., Bashton, M., Gough, J., and Teichmann, S. A.(2004) Supra-domains: evolutionary units larger than single protein do-mains. J. Mol. Biol. 336, 809 – 823

45. Orengo, C. A., and Thornton, J. M. (2005) Protein families and their evo-lution-a structural perspective. Annu. Rev. Biochem. 74, 867–900

46. Deans, A. J., and West, S. C. (2009) FANCM connects the genome insta-bility disorders Bloom’s syndrome and Fanconi anemia. Mol. Cell 36,943–953

47. Singh, T. R., Saro, D., Ali, A. M., Zheng, X. F., Du, C. H., Killen, M. W.,Sachpatzidis, A., Wahengbam, K., Pierce, A. J., Xiong, Y., Sung, P., andMeetei, A. R. (2010) MHF1-MHF2, a histone-fold-containing proteincomplex, participates in the Fanconi anemia pathway via FANCM. Mol.Cell 37, 879 – 886

48. Yan, Z., Delannoy, M., Ling, C., Daee, D., Osman, F., Muniandy, P. A.,Shen, X., Oostra, A. B., Du, H., Steltenpool, J., Lin, T., Schuster, B.,Decaillet, C., Stasiak, A., Stasiak, A. Z., Stone, S., Hoatlin, M. E., Schin-dler, D., Woodcock, C. L., Joenje, H., Sen, R., de Winter, J. P., Li, L.,Seidman, M. M., Whitby, M. C., Myung, K., Constantinou, A., andWang, W. (2010) A histone-fold complex and FANCM form a con-served DNA-remodeling complex to maintain genome stability. Mol.Cell 37, 865– 878

49. Tao, Y., Jin, C., Li, X., Qi, S., Chu, L., Niu, L., Yao, X., and Teng, M. (2012)The structure of the FANCM-MHF complex reveals physical features forfunctional assembly. Nat. Commun. 3, 782

50. Pluchon, P. F., Fouqueau, T., Creze, C., Laurent, S., Briffotaux, J., Hogrel,




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

G., Palud, A., Henneke, G., Godfroy, A., Hausner, W., Thomm, M., Nico-las, J., and Flament, D. (2013) An extended network of genomic mainte-nance in the archaeon Pyrococcus abyssi highlights unexpected associa-tions between eucaryotic homologs. PLoS One 8, e79707

51. Ulrich, H. D., and Takahashi, T. (2013) Readers of PCNA modifications.Chromosoma 122, 259 –274

52. Li, Z., Pan, M., Santangelo, T. J., Chemnitz, W., Yuan, W., Edwards, J. L.,

Hurwitz, J., Reeve, J. N., and Kelman, Z. (2011) A novel DNA nuclease isstimulated by association with the GINS complex. Nucleic Acids Res. 39,6114 – 6123

53. Yuan, H., Liu, X. P., Han, Z., Allers, T., Hou, J. L., and Liu, J. H. (2013)RecJ-like protein from Pyrococcus furiosus has 3�-5� exonuclease activityon RNA: implications for proofreading of 3�-mismatched RNA primers inDNA replication. Nucleic Acids Res. 41, 5817–5826




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Yoshizumi IshinoShyogo Sugiyama, Toshio Ando, Natsuko Goda, Takeshi Tenno, Hidekazu Hiroaki and Sonoko Ishino, Takeshi Yamagami, Makoto Kitamura, Noriyuki Kodera, Tetsuya Mori,

and Nuclease Domains of the Archaeal Hef ProteinMultiple Interactions of the Intrinsically Disordered Region between the Helicase

doi: 10.1074/jbc.M114.554998 originally published online June 19, 20142014, 289:21627-21639.J. Biol. Chem.

10.1074/jbc.M114.554998Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2014/06/19/M114.554998.DC1

http://www.jbc.org/content/289/31/21627.full.html#ref-list-1

This article cites 53 references, 9 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M114.554998

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;289/31/21627&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/289/31/21627

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=289/31/21627&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/289/31/21627

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2014/06/19/M114.554998.DC1

http://www.jbc.org/content/289/31/21627.full.html#ref-list-1

http://www.jbc.org/

Multiple Interactions of the Intrinsically Disordered Region between ...

Documents

Transcript of Multiple Interactions of the Intrinsically Disordered Region between ...