Aggregative organization enhances the DNA end-joining process that is mediated by DNA-dependent protein kinase Masahiko Takahagi and Kouichi Tatsumi Research Center for Radiation Safety, National Institute of Radiological Sciences, Chiba, Japan Keywords coacervate; DNA aggregation; DNA endjoining; DNA-PK; S ⁄ MAR Correspondence K Tatsumi, Research Center for Radiation Safety, National Institute of Radiological Sciences, 9–1, Anagawa 4, Inage-ku, Chiba 263–8555, Japan Fax: +81 43 255 6497 Tel: +81 43 206 3087 E-mail: tatsumi@rea.or.jp (Received April 2006, accepted 11 May 2006) doi:10.1111/j.1742-4658.2006.05317.x The occurrence of DNA double-strand breaks in the nucleus provokes in its structural organization a large-scale alteration whose molecular basis is still mostly unclear Here, we show that double-strand breaks trigger preferential assembly of nucleoproteins in human cellular fractions and that they mediate the separation of large protein–DNA aggregates from aqueous solution The interaction among the aggregative nucleoproteins presents a dynamic condition that allows the effective interaction of nucleoproteins with external molecules like free ATP and facilitates intrinsic DNA end-joining activity This aggregative organization is functionally coacervate-like The key component is DNA-dependent protein kinase (DNA-PK), which can be characterized as a DNA-specific aggregation factor as well as a nuclear scaffold ⁄ matrix-interactive factor In the context of aggregation, the kinase activity of DNA-PK is essential for efficient DNA end-joining The massive and functional concentration of nucleoproteins on DNA in vitro may represent a possible status of nuclear dynamics in vivo, which probably includes the DNA-PK-dependent response to multiple double-strand breaks DNA double-strand breaks (DSBs) are a serious threat to the genetic integrity of organisms, causing cell death if not repaired The repair mechanism for DSBs resides not only in catalytic processes but also in the association with chromatin structures [1,2], although the details in the higher-order context remain obscure Some evidence has implicated structural alterations in the vicinity of DSB sites DSBs can form nuclear foci linked to phosphorylated histone H2AX (c-H2AX) [3], which is responsible for the redistribution of repair factors to DSB sites [4], although it is dispensable for initial damage recognition [5] Approximately 2000 c-H2AX molecules accumulate per focus in a normal human cell, suggesting reorganization of chromosomal DNA over a region of Mbp order [6] c-H2AX-associated foci are morphologically dynamic; the DSB-containing chromosome domains can be mobile, and in a subpopulation of damaged cells, they can juxtapose via an adhesion process irrespective of DNA repair processes [7] In addition to these large-scale responses to DSBs, real-time analysis of the temporo-spatial distribution of DNA repair factors in situ in living cells has been providing us with striking information For instance, even following exposure to ionizing radiation (IR), the DNA end-binding factor Ku moves rapidly throughout the nucleus but associates transiently with filamentous nuclear substrates [8] A checkpoint regulator NBS1, the product of the Nijmegen breakage syndrome gene, shuttles rapidly between DSB sites and the flanking chromatin [9] These findings indicate that the action of DSB-interactive proteins within the nuclear microenvironment must be coupled with the mobile state of those proteins Abbreviations DSB, double-strand break; DNA-PK, DNA-dependent protein kinase; EJ, end-joining; c-H2AX, phosphorylated histone H2AX; HMGB, high mobility group box; IR, ionizing radiation; LD, linear duplex; NC, nicked circular; NF, nuclear fraction; SC, supercoiled; S ⁄ MAR, scaffold ⁄ matrix attached region; ss, single strand; SSC, single-strand circular FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3063 Aggregation-coupled DNA end-joining M Takahagi and K Tatsumi One hallmark of the initial response to DSBs is the rapid activation of ATM (product of the ataxia telangiectasia mutated gene), the protein kinase that is mutated in the human hereditary disease ataxia telangiectasia: induced chromatin alterations relay its intermolecular autophosphorylation and dimer dissociation [10] Over 50% of the ATM molecules in primary fibroblast-like cells are activated within after exposure to 0.5 Gy of IR This activation is triggered even by treatment with hypotonic conditions or chemical chromatin modifiers Nevertheless, detectable DSBs are not induced by such treatments It seems likely therefore that chromatin alterations, at a certain distance from DSB sites, initiate the ATM activation Such distant effects of IR have been similarly recognized as a rapid change of the topological constraints on chromosomal DNA in the nuclei [11–13] The magnitude of these nuclear changes appears to be relevant to the anchoring of DNA loop domains on the nuclear scaffold or matrix, where the nuclear scaffold ⁄ matrix attached region (S ⁄ MAR) DNA sequences are thought to mediate the higher-order networking of S ⁄ MAR-binding proteins This property has been biochemically defined as the preferential aggregation of those proteins onto specific DNA The potential for aggregation has been characterized for major nucleoproteins including histone H1 [14], topoisomerase II [15], lamin B1 [16] and SAF-A (hnRNP U) [17] It is known that the aggregation of DNA is compatible with the functions of proteins under in vitro conditions Polyamine-mediated aggregation involves the suitable arrangement of DNA helices, allowing the efficient catenation of the circular duplexes by Escherichia coli topoisomerases, the free exchange of DNA between neighboring aggregates by the enzyme [18], and the activation of transcription by E coli RNA polymerase [19] The E coli recombination factor RecA promotes coaggregation between single- and double-strand DNA, and the diffusible nature of components within the nucleoprotein networks encourages inefficient homologous pairing [20,21] These examples illustrate that an aggregative organization underlies the interplay among proteins and DNA and that it serves with a compact but fluid state for catalytic reactions Thus, in view of the large-scale nuclear responses in human cells, it is proposed that aberrant DNA structures, particularly DSBs, are recognized and processed through interactions with repair factors and S ⁄ MARbinding proteins in an aggregative context We examined under in vitro conditions whether DSBs mediate the aggregative assembly of nucleoproteins that assures 3064 reliable repair reactions Here we show that an extensive aggregation of human nucleoproteins, including DNA repair kinase DNA-PK, is mobilized onto DNA with ends in a cell-free system, and that it produces the characteristic ‘coacervate-like organization’, which greatly stimulates DNA end-joining Results and Discussion Aberrant types of DNA coaggregate with nucleoproteins To know whether the assembly of nucleoproteins with DSBs is aggregative, we examined the precipitable property of the complex The assay system adopted not only linear duplex as a model substrate of DSBs but also other forms of DNA including supercoiled, nicked circular and single-strand circular DNA as competitors The reaction mixtures containing nuclear fractions (Fig 1A) from the human lymphoblastoid cells, WI-L2-NS, were centrifuged after incubation with DNA in solution A visibly gelated protein–DNA complex precipitated at the bottom of the test tubes After being separated from the supernatant and deproteinized, the precipitate was subjected to gel electrophoresis for size-based analysis of DNA Except for the supercoiled form, all aberrant types were preferential targets for aggregation, which occurred depending on the concentration of protein added (Fig 1B, upper panel, lanes 3–7) Under the physiological ionic strength (0.15 m NaCl), the formation of stable aggregates required a large amount of extracts To avoid this problem and the complexity between substrates, the aggregation of a linear DNA was examined under a lower ionic condition (0.05 m NaCl), and the extent of the aggregation was compared to residual molecules in the supernatant (Fig 1C, upper panel) The increasing amounts of extracts quantitatively precipitated the DNA, ending up with saturation at a full extent (lanes and 14) Considering the biological significance of the aggregation, we expected that the aggregates would be structurally fluid and anisotropic like the polyamine-induced DNA aggregation and so would possess the feature of coacervate [22], defined as a dynamic phase which appears by self-assembly of colloidal biopolymers in dilute solutions [23] Nucleoprotein aggregation enhances intrinsic DNA end-joining activity With regard to the biochemical aspects of the aggregation, we examined whether or not the assembly is related to DNA repair processes We adopted as a simple FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS B DEAE Sepharose 0.5 M / FT No DNA Human Nuclear Extract Substrate A Aggregation-coupled DNA end-joining Input DNA M Takahagi and K Tatsumi - 1.0 M NaCl Nuclear fraction Denatured DNA Cellulose NC Nuclear SSC fraction SC * * LD 0.2 M / FT 0.5 M 1.0 M NaCl DNA affinity fraction DNA affinity fraction Mono Q 0.1 M / FT 0.5 M 1.0 M NaCl Mono Q fraction Mono Q fraction Active fraction Input C ppt - sup - Nuclear fraction LD DNA affinity fraction LD Mono Q fraction LD 10 11 12 13 14 15 Fig Purification of DNA aggregation activity from human cells (A) Scheme for the purification of DNA aggregation activity (B) DNA aggregation activity in chromatographic fractions Nucleoproteins from WI-L2-NS cells were sequentially separated through DEAE sepharose (nuclear fraction), denatured DNA cellulose (DNA affinity fraction) and Mono Q columns (Mono Q fraction) Input DNA contained four forms of uX174 phage DNA, including 0.05 lg of HaeIII-digested linear duplex (LD), 0.05 lg of single-stranded circular (SSC) and 0.05 lg of supercoiled (SC) DNA containing nicked (NC) forms Input DNA was incubated with each fraction increasing in amount by two-fold at physiological ionic strength (150 mM NaCl) Either 240 lg of nuclear fraction, 70 lg of DNA affinity fraction or 70 lg of Mono Q fraction was used as a maximal protein concentration (lane 7) After centrifugation of reaction mixtures, precipitated DNA was analyzed by 1% agarose gel electrophoresis DNA was visualized by staining with ethidium bromide Results are represented by negative images Asterisks indicate nucleic acids derived from cells (C) Coaggregation of linear DNA by separated fractions Reactions of fractions with EcoRI-digested pUC18 DNA (0.1 lg) as an input substrate (LD) were conducted in the presence of 50 mM NaCl After centrifugation of reaction mixtures, both centrifugal precipitates (ppt.) and its supernatant (sup.) were resolved on an agarose gel Maximal aggregation of DNA was achieved by using either 5.4 lg of nuclear fraction, 1.0 lg of DNA affinity fraction or 0.7 lg of Mono Q bound fraction (lanes and 14) model system the rejoining reaction of DNA ends (end-joining; EJ), which has been characterized as the primary repair mechanism of DSBs [1] EJ reaction was carried out under a condition for full aggregation of linear DNA (0.1 lg) (Fig 1C, lanes and 14) After a 2.7 kbp linearized-plasmid DNA was coaggregated with human nuclear fractions, an EJ reaction was initiated by the addition of ATP as a cofactor into the surrounding buffer solution (Fig 2A) At low ionic strength (50 mm NaCl), the intermolecular joining of DNA ends was greatly enhanced This stimulation was also noted at higher ionic strength (150 mm NaCl), although the yield of EJ products was lower in accordance with the efficiency of DNA aggregation (unpublished work) Even when the precipitable aggregate was detached from the bottoms of the tubes, the EJ FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3065 Aggregation-coupled DNA end-joining Input A In aggregate M Takahagi and K Tatsumi In solution 10 30 60 120 10 30 60 120 Time (min) Multimer Trimer Dimer Monomer 43 57 68 72 73 B 16 20 25 30 35 5'-overhang 3'-overhang - + + + + + + + + + - - + + + + + + + + + - EJ products (%) blunt - - + - + ++ + + + + + - DNA affinity fraction Fraction Aggregation T4 DNA ligase ATP Multimer DimerTrimer Monomer 10 11 12 13 14 15 Mono Q fraction Multimer DimerTrimer Monomer 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Fig End-joining activity is associated with nucleoprotein-DNA coaggregation Linearized pUC18 DNA of an input substrate (Monomer) was treated with separated fractions under conditions for its maximum aggregation as shown in Fig 1C EJ reaction was initiated by the addition of ATP onto the aggregates (A) Nucleoprotein–DNA coaggregation enhances intrinsic EJ activities in nuclear fraction EJ reactions were monitored as a function of incubation time and were compared between in aggregate and in solution The EJ occurred predominantly in an intermolecular manner, producing dimer, trimer and multimers (B) Preference of DNA end-structure in the aggregation-coupled EJ Three types of linearized pUC18 (0.1 lg), 5¢-overhangs, 3¢-overhangs and blunt ends, were incubated with DNA affinity (1.0 lg) and Mono Q bound (0.7 lg) fractions Under identical conditions except for centrifugal aggregation, T4 DNA ligase promoted not only intermolecular but also intramolecular ligation, parts of which were resolved with faster mobility than the monomer (lanes 2, 7, 17 and 22) activity was not enhanced (supplementary Fig S1) The activation can therefore be attributed to the aggregative organization of nucleoproteins but not to contact with the plastic surface A similar activity was detectable in nuclear fractions from several other human lymphoblastoid cell lines, HeLa S3 cells and placenta tissue (unpublished work), implying that the promotion of EJ by aggregation is common to many types of human cells It is very likely that the resultant organization of nuclear components from dilute aqueous solutions involves either preferential assembly of DSB repair factors containing DNA ligase activity, valid synapsis of DNA ends, or sequestration from 3066 inhibitory factors Because the aggregates with efficient EJ capacity are selective in the intermolecular mode of EJ and are readily accessible to external ATP molecules, the internal structure must be fluid and must be an open system Thus, we argue that this organization of nucleoproteins on linear DNA qualifies as coacervation, which has been implicated in the mechanism for the condensed process of prebiotic charged polymers and has been established as a condition interactive with the external environment during their continuous synthesis and breakdown [23, and references therein] DNA aggregation activity is coupled with EJ activity To further confirm the relationship between protein– DNA aggregation and EJ activity, we fractionated the flow-through of DEAE-Sepharose (the nuclear fraction) by sequential chromatography with denatured-DNA coupled affinity columns and Mono Q anion-exchange columns (Fig 1A) Both the aggregation activity and EJ activity were enriched by a series of fractionations As is the case with nuclear fractions, the DNA affinity- and Mono Q bound fractions coaggregated with linear, nicked and singlestrand DNA with a similar preference (Fig 1B) On the other hand, the substrate specificity of EJ was distinct among fractions (Fig 2B) For both DNA affinity- and Mono Q bound fractions, 5¢-overhangs were the most reactive form, and 3¢-overhangs were also relatively reactive The reactivity of blunt ends was much lower in DNA affinity fractions, but was as high as that of 3¢-overhangs in Mono Q fractions (compare lanes 24 and 29) The active fraction from the Mono Q column was also devoid of ATP-independent EJ activity, which was present in DNA affinity fractions (lanes 5, 10 and 15) These results indicate that EJ and aggregation activities can be closely associated but are functionally independent The difference in protein composition among fractions may account for their distinct reactivity with endstructures in EJ (Fig 3A, compare lanes and 4) Apparently, our purification scheme is different from the established one for major end-joining activity [24] In addition, immunoblot analysis showed that the specific ligase complex, including DNA ligase IV and XRCC4 proteins, fully passed through the DNA affinity-column, so they were not present in the Mono Qactive fraction (supplementary Fig S2) We reason that the EJ activity enriched here seems to depend on some other ligase activities The candidate could be DNA ligase III, which has recently been demonstrated FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS M Takahagi and K Tatsumi Identification of proteins aggregative with aberrant DNA NC SSC Aggregative proteins LD DNA affinity Mono Q Marker NF Separated fractions no DNA relaxed A Aggregation-coupled DNA end-joining (kDa) 225 150 100 75 DNA-PKcs 180 k 170 k SAF-A 98 k Nucleolin Ku80 Nucleolin∆ Ku70 50 35 Histone H1 28 k HMGB1 HMGB2 25 Marker HMGB2 HMGB1 B DNA-PKcs Ku70/Ku80 SAF-A Nucleolin Nucleolin∆ Histone H1 Marker Fig Identification and isolation of the aggregative nucleoproteins (A) Nucleoproteins in chromatographic fractions and in aggregates with DNA Proteins were analyzed by SDS ⁄ PAGE in a 10% gel and stained with Coomassie brilliant blue The profile of fractionated proteins included nuclear fraction (NF; lane 2), DNA affinity fraction (lane 3) and Mono Q fraction (lane 4) For aggregation of proteins, the Mono Q bound fraction (7.0 lg) was incubated with no DNA (lane 5) or aberrant types of DNA (1.0 lg), including relaxed closed circular (lane 6), NC, nicked form (lane 8) and SSC, single-strand circular (lane 9) derived from uX174, and linear duplex (LD) of EcoRIdigested pUC18 (lane 7) Several proteins were identified by amino acid sequencing Relatively abundant proteins are marked with arrows (B) Major nucleoproteins isolated from Mono Q fraction Isolated preparations of DNA-PKcs (2.0 lg), Ku70 ⁄ Ku80 (2.0 lg), SAF-A (0.4 lg), nucleolin (0.8 lg), nucleolinD (0.4 lg) and histone H1 (2 lg) were resolved by a 10% SDS ⁄ PAGE Purified HMGB1 (2 lg) and HMGB2 (2 lg) were analyzed on a 12% SDS ⁄ PAGE The sizes of marker proteins are 94, 67, 43, 30 and 20 k in the context of an alternative EJ pathway [25,26] The characterization of ligase III is underway in the aggregation system The aggregative nucleoproteins were collected under conditions that allow maximal precipitation of input DNA including linear, nicked and single-strand (ss) forms (Fig 3A, lanes 7–9), and were resolved on an SDS ⁄ PAGE For identification, the primary structure of major polypeptides was determined by peptide mapping and partial protein sequencing When Mono Q bound fractions were subjected to assembly with DNA, all aggregates contained proteins of DNA-PKcs, 180 k, 170 k, SAF-A, 98 k, nucleolin, Ku80, Ku70, histone H1 and 28 k proteins, regardless of the type of damage (Fig 3A, lanes 7–9) In the aggregation of linear and nicked DNA, and high mobility group box (HMGB)1 and HMGB2 proteins were additional elements (lanes and 8) ssDNA induced a noticeable accumulation of 95 k and nucleolinD proteins in addition to DNA-PKcs, Ku80 and Ku70 (lane 9) On the other hand, a relaxed form of closed circular doublestrand DNA as an undamaged substrate limited the composition and quantity of aggregative proteins (lane 6) It should be noted that proteins that specifically aggregate with linear DNA were similar to those that aggregate with nicked DNA, and that some of them also assemble with ssDNA Interestingly, the components of DNA-dependent protein kinase, DNA-PKcs, Ku80 and Ku70, were the common factors to aggregate with preference to damaged types By contrast, histone H1 was detectable in the aggregates with all kinds of DNA examined Also, these results indicate that centrifugal precipitation was a uniquely effective means to collect and classify the factors aggregative with aberrant DNA Characterization of DNA aggregation-promoting factors We expected that the candidates that coaggregate with aberrant DNA would be relatively abundant factors Major proteins in the Mono Q bound fraction, including DNA-PKcs, SAF-A, nucleolin, nucleolinD, Ku (as a Ku70 ⁄ Ku80 heterodimer), HMGB1 and HMGB2, have been purified by column chromatography (Fig 3B) In contrast with experiments using partially purified fractions, none of the isolated proteins coprecipitated with DNA at physiological ionic strength However, as the NaCl concentration in reaction mixtures was decreased from 150 mm to 50 mm, three proteins, DNA-PKcs, SAF-A and nucleolin, began to interact selectively with specific targets in an aggregative manner (Fig 4) Nucleolin assembled FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3067 +Mg - -Mg - M Takahagi and K Tatsumi Input Input Aggregation-coupled DNA end-joining +Mg - -Mg Nucleolin∆ Nucleolin DNA-PKcs SAF-A Ku70/Ku80 HMGB1 HMGB2 NC SSC Histone H1 SC LD preferentially on ssDNA but not on any variations of duplex DNA This interaction required Mg2+ as a cofactor Interestingly, nucleolinD, whose N-terminal 138 residues of 709 amino acids are spontaneously truncated in most human lymphoblastoid cell lines (unpublished work), was deficient in aggregation activity, implying that the truncated region may be indispensable for aggregation Likewise, DNA-PKcs promoted predominant aggregation of ssDNA when incubated at relatively low concentrations At higher ratios of protein to DNA, DNA-PKcs started to interact with nicked and linear DNA but not with supercoiled DNA Again, Mg2+ moderately facilitated protein–DNA interaction SAF-A coaggregated with all damaged types of DNA in a Mg2+-dependent manner This preference is conditional since supercoiled DNA was aggregative in the absence of Mg2+ These results suggest that the DNA repair factor DNA-PKcs is an aggregation factor that responds to aberrant DNA structures in a manner similar to that of S ⁄ MAR-binding proteins, including nucleolin and SAF-A None of the isolated Ku, HMGB1 or HMGB2 coaggregated with any DNA, although in Mono Q fractions they are aggregative dependent on linear and nicked DNAs (Fig 3A, lanes and 8) Conversely, histone H1 aggregated with all substrates including supercoiled DNA, a finding consistent with the fact that it was accompanied in protein analysis by all forms (Fig 3A, lanes 6–9) The basic polypeptide is thought to interact electrostatically with negativecharged DNAs in a nonspecific manner as well as multivalent cations of polyamines [27] This way may 3068 Fig Selective aggregation of different forms of DNA by purified proteins Conditions were similar to those used in Fig 1C Protein-mediated aggregation of DNA substrates (0.15 lg) that contained linear duplex (LD), nicked form (NC), single-strand circular (SSC) and supercoiled (SC) were examined in the absence or presence of 10 mM Mg2+ The two-fold increasing concentrations of nucleolin (0.12 lg), nucleolinD (0.12 lg), DNA-PKcs (0.3 lg), SAF-A (0.15 lg), Ku (0.3 lg), HMGB1 (0.12 lg), HMGB2 (0.12 lg), and histone H1 (0.12 lg) proteins were subjected to experiments be distinct from that of nonbasic factors of DNA-PKcs, SAF-A and nucleolin, which were stably concentrated on anion-exchange resins (Experimental procedures) With regard to the action mode, the role of the N-terminal portion of nucleolin is suggestive This region includes the basic and repeated octapeptides that are similar to histone H1 [28] and acidic stretches that are able to be associated with histone H1 [29] The truncation of the former motif and a part of the latter one abolished the ssDNA-specific aggregation activity (Fig 4) but remained a valid binding capacity to ssDNA (unpublished work) Therefore, the aggregation is likely to occur not only by DNA binding but also through an aggregation-promoting process, probably DNA-mediated intra- or intermolecular association We presume that the structural and functional distribution on relatively large proteins may be involved in the aggregation of aberrant DNAs DNA-PK is the key factor for aggregation-coupled EJ The functional interplay between DNA-PKcs and Ku in the aggregates is noteworthy because they function in a common pathway, and the protein kinase activity of DNA-PKcs is greatly stimulated by Ku during assembly at DNA ends [30] We examined whether the aggregation activity of DNA-PKcs is regulated by Ku While DNA-PKcs interacts exclusively with nicks and ssDNA, the addition of increasing amounts of Ku stimulated progressive aggregation of linear DNA without size-dependency in the range from 200 to 1350 FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS Aggregation-coupled DNA end-joining Input No protein M Takahagi and K Tatsumi A Fig Activity of DNA-PKcs in the aggregates (A) Ku facilitates DNA-PKcs-mediated DNA aggregation Ku (0.15, 0.3, 0.6, 1.2 lg) was added to the mixture of DNA-PKcs (1.2 lg) or SAF-A (0.6 lg) with DNA (0.15 lg) including LD, NC, SSC and SC The effect of Ku on DNA-PKcs-DNA coaggregation is indicated by a bracket (upper panel) (B) Aggregation is compatible with the kinase activity of DNA-PKcs The combination of DNA-PKcs (1.6 lg) and Ku (1.0 lg) or Mono Q fraction (1.4 lg) was coaggregated with linearized pUC18 in the absence or presence of wortmannin Phosphorylation was initiated by the addition of c-P32-labeled ATP to the aggregates, and the radioactive proteins were resolved on an SDS ⁄ PAGE (C) The kinase activity of DNAPKcs regulates the aggregation-coupled EJ The effect of the addition of purified DNA-PKcs (0.2, 0.4, 0.8, 1.6 lg) to wortmannin-inactivated aggregates was also tested (lanes 6–9) The aggregate and the supernatant (sup.) were analyzed separately (lanes 10 and 11) The conditions in lanes 4, and are parallel to those in lanes 3–5 in Fig 5B, respectively B DNA-PKcs Ku70/Ku80 Mono Q fraction Wortmannin Ku70/Ku80 (kDa) 225 150 100 75 DNA-PKcs DNA-PKcs Ku80 Ku70 50 35 SAF-A 25 C Wortmannin DNA-PKcs ATP Aggregate after sup incubation Multimer Trimer Dimer Monomer 10 11 base pairs (Fig 5A, upper panel) By contrast, Ku failed to promote the assembly of SAF-A on linear DNA under conditions similar to those optimal for selective aggregation of nicks and ssDNA (Fig 5A, lower panel) The effect of Ku on aggregating activities of other purified factors was negative as well These data indicate that through protein–protein interaction, Ku operates to specifically promote the aggregation of DNA-PKcs on DNA ends The previous gel-shift study indicated that highly purified DNA-PKcs binds to linear DNA in the absence of Ku and predominantly forms a large complex without any mobility shift in the well of the gel [31] The stability of the complex against salt is also enhanced by the presence of Ku Interestingly, the complex formation was easily inhibited by ssDNA, including poly(dT) and hairpin-ended DNA, although these are unable to activate DNA-PK kinase efficiently [31,32] In addition, since such an immobile complex in the well was found under conditions for aggregation between DNA-PKcs and DNA, we conclude that the complex in this study has aspects of an aggregate Electron microscopic imaging has consistently shown that DNA-PK is able to link multiple linear DNA fragments into a large complex [33] The autophosphorylation of DNA-PK causes the disruption of such a complex formation, which is biochemically dynamic If so, our centrifugal manipulation appears to conduct higher-order accumulation of preexisting similar complexes in solution and could generate a preferential condition for the intermolecular DNA end-joining The autophosphorylation of DNA-PKcs was detectable in aggregates with linear DNA (Fig 5B, lane 1) as well as in reactions in solution Ku also enhanced the autophosphorylation of DNA-PKcs in addition to its own phosphorylation (Fig 5B, lane 2) This biochemical evidence supports the view that the DNAPKcs-mediated aggregates stay in a catalytically competent state A further concern is whether the protein kinase activity of DNA-PKcs is really associated with the EJ reaction in the context of protein–DNA coaggregative networks To answer this question, we examined the effect of the DNA-PK inhibitor wortmannin [34] This inhibitor covalently binds to the catalytic domain on DNA-PKcs [35], thereby leading to an abortion of the EJ process in solution through an irreversible FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3069 Aggregation-coupled DNA end-joining M Takahagi and K Tatsumi inactivation of kinase activity [36,37] Preincubation with the inhibitor did not alter aggregation activity in Mono Q bound fractions (Fig 5C, compare lanes and 3) However, the resultant aggregates yielded only a slight amount of EJ product in the presence of ATP (Fig 5C, lane 5) The subsequent addition of purified DNA-PKcs to the inactivated aggregate restored EJ activity to a considerable degree, depending on the amount of DNA-PKcs added (Fig 5C, lanes 6–9) Here DNA-PK activity in the aggregates showed a wortmannin-sensitivity and was restored by exogenous DNA-PKcs (Fig 5B, lanes 3–5) Since the aggregate stuck to the bottom of the test tube during reactions (Fig 5C, compare lanes 10 and 11), DNAPKcs molecules supplied to the supernatant must be interactive with the aggregate and must be accessible to its critical sites Thus, the aggregate has the quality of an open system under continuous interaction with the surrounding environment Taken together, we can describe the DNA end-directed aggregation as a coacervate-like organization, which serves as an activated state to enable DNA-PK-dependent end-joining This term, coacervate-like organisation, means a molecular network that has properties of naturally occurring coacervate but is obtainable through centrifugal manipulation in an in vitro system Additionally, when a DNA-PKcs-enriched subfraction, which was isolated from other aggregative proteins including Ku through a Mono Q column (refer to Experimental procedures), a significant portion of EJ activity was co-eluted (supplementary Fig S3) The activity was stimulated by aggregation and was wortmannin-sensitive in the presence of purified Ku The evidence from this reconstruction system emphasizes that DNA-PKcs and Ku are key elements for the aggregation-coupled EJ Aggregative organization as a large-scale response to DNA damage To our surprise, the DNA repair factors DNA-PKcs and Ku themselves participate in aggregation, supporting the tight link of aggregative organization with repair processes Under reconstituted conditions in vitro, nick (single-strand break) and ssDNA (unpaired form) lesions, which could release the superhelical torsion of DNA loops are also mediators for aggregation, and are even predominant substrates for the aggregative action of DNA-PKcs Experiments have shown that cellular protein fractions can target those lesions with affinity similar to that of DNA ends We have also noted that the aggregates prepared from one type of DNA lesion assemble more readily with 3070 DNA of other types of lesions diffused in external solution under physiological conditions (unpublished work) This predisposition implies the possibility that in vivo, instead of centrifugal force, the pre-existing aggregates may develop into damage-mediated ones It is presumable that such nucleation of aggregates takes place at AT-rich S ⁄ MAR DNA of an anchor of chromatin loops that tend to transform into ssDNA by unwinding [38] and are potential sites for aggregation of S ⁄ MAR-binding proteins In this view, the direct association between DNA-PK and S ⁄ MAR-DNA or -binding proteins is a considerable point An experimental observation seems to be the case that in an in vitro system, the binding of DNA-PK to DNA ends is prerequisite for the secondary assembly with S ⁄ MAR DNA concomitant with its interactive factors including SAF-A [39] It should be also noted that more frequent nucleation may originate from reversible association between nucleosomal units of chromatin in equilibrium of ionic conditions [40,41] Probably, the extent of the aggregation would be regulated by DNA damage-responsive chromatin factors, and one of the candidates may be linker histone H1 of an aggregative protein IR provokes ATM-dependent dephosphorylation of histone H1 [42] A recent study demonstrates that whereas histone H1 is an inhibitory factor for EJ reaction, its phosphorylation by DNA-PK results in the promotion of EJ even in the presence of nucleosomes [43] We presume that in addition to the nucleosome component H2AX, the change in the phosphorylated state of histone H1 would regulate the aggregative nature of the array of nucleosomes Concerning DNA-PKcs, the results from the previous in vitro experiments can be applied to the action mode according to a simple stoichiometry: one molecule of DNA-PKcs is recruited to each Ku heterodimer-bound DNA end to form an active kinase complex, and synapsis of the complexes promotes the juxtaposition between the opposing DNA ends, prior to their processing and ligation [44] However, this scheme cannot fully explain the evidence for the in vivo modification and localization of DNA-PKcs An exposure of cells to IR induces the autophosphorylation of many DNA-PKcs molecules at DSB sites [45], which are coupled with DSB-induced c-H2AX nuclear foci and are likely to be concomitant with the accumulation of activated ATM molecules [46] Such accumulation of the phosphorylated DNA-PKcs into the foci may be parallel to its quantitative requirement for in vitro aggregation; the conditional ratio of DNA-PKcs in aggregates is calculated as being >10 molecules per a single end of 2.7 kbp linear DNA Very likely, the analysis of components of FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS M Takahagi and K Tatsumi Mono Q fraction coaggregated with linear DNA (Fig 3A, lane 7) indicates that one DNA end attracts a similarly excessive number of molecules of nucleoproteins Our estimation, however, remains tentative until their distribution on individual DNA is determined Interestingly, rodent cells that are defective in Ku80 or DNA-PKcs are incomplete in the repair of chromatin loops containing multiple DSBs, which were induced by high doses (> 50 Gy) of X-ray or c-ray [47,48] DNA-PK components are likely to be involved in two distinct processes: fast kinetics for single DSB and a slow one for multiple DSBs in loops The way that deals with multiple DSBs may be comparable to our present study Presumably, multiple damaged sites localized within a chromatin region may elicit another pathway dependent on DNA-PK Alternatively, an experiment using a radiation, which was irradiated at 0.5 Gy-equivalent dose per one track, demonstrated an interesting behavior of multiple DSBs [7] The introduction of a-particle track onto the cellular nucleus provokes either rapid mobility or adhesion of the DSB-induced c-H2AX nuclear foci This appears to be representative of the interaction between multiple DSBs on different chromosomes, which is likely to predispose to an irregular rejoining Given high fluidity and self-organization, one speculation is that the structural essence of the nuclear foci could be illustrated by the coacervate-like nature of the DSB-directed aggregates Concluding remarks The present results demonstrate that the aggregative organization composed of human nucleoproteins and aberrant DNA is structurally flexible and biochemically active, suggesting that the coacervate-like phase transition from random diffusion to ordered network could be an architectural strategy in DNA damageresponse The organization feasible for catalytic repair reactions has several remarkable characteristics: (i) the recognition and incorporation of multiple damage sites in a noncatalytic fashion; (ii) the concentration of factors diluted in solution; (iii) the acceleration of reactions; and (iv) the involvement of DNA aggregation-promoting factors containing S ⁄ MAR-binding proteins Practically, the aggregation-based technique is versatile for extraction of DNA-interactive factors and for their higher-order assembly, and will contribute to a thorough understanding of the mechanism for huge accumulation, high mobility, transient interaction and activity-regulation of proteins in the nuclear environment Aggregation-coupled DNA end-joining Experimental procedures Cells The human lymphoblastoid cells, WI-L2-NS (CRL8155), were provided from the Health Science Research Resources Bank (HSRRB), Osaka, Japan (formerly the Japanese Cancer Research Resources Bank) The cells were cultured in RPMI1640 medium supplemented with 10% fetal calf serum, and were incubated at 37 °C in a humidified atmosphere with 5% CO2 DNA substrates DNA substrates for assays were ss circular, supercoiled, nicked, and HaeIII-digested linear DNA derived from uX174 phage (New England BioLabs, Beverly, MA, USA) The relaxed closed circular form was made by sealing nicked DNA with T4 DNA ligase (TaKaRa, Shiga, Japan) The linearized forms of pUC18 plasmid DNA were prepared by digestion with the restriction enzymes EcoRI, PstI and ScaI (TaKaRa) as alternative substrates DNA concentrations were determined from absorbance at 260 nm Cellular fractions and proteins Nuclear extracts were prepared as described previously with the following modifications [49] Cells (4 · 109 cells at a density of · 106 cellsỈmL)1) were treated by swelling in three packed volumes of A buffer [10 mm Hepes–KOH, pH 7.9, 1.5 mm MgCl2, 10 mm KCl, 0.5 mm dithiothreitol, and 0.5 mm phenylmethanesulfonyl fluoride] on ice for 15 The swollen cells were homogenized by 20 strokes of a B-pestle in a Dounce homogenizer (Wheaton, Millville, NJ, USA) The nuclei collected by centrifugation at 2500 g for 10 were rinsed and suspended in B buffer (20 mm Hepes–KOH, pH 7.9, 0.5 m NaCl, 25% glycerol, 1.5 mm MgCl2, 0.2 mm EDTA, 0.5 mm dithiothreitol, 0.5 mm phenylmethanesulfonyl fluoride) Following holding on ice for 30 min, the nuclear extract was obtained as the supernatant after centrifugation at 2500 g For chromatographic procedures, the nuclear extract was filtered through a 0.20 lm pore membrane (Minisart plus lter; Sartorius, Gottingen, Germany) and then applied to ă a column with DEAE Sepharose resin (Amersham Biosciences, Piscataway, NJ, USA) in R buffer (20 mm Tris ⁄ HCl, pH 7.5, 0.5 mm EDTA, mm 2-mercaptoethanol) with 0.5 m NaCl The flow-through fraction was referred to as the nuclear fraction The fraction was diluted to 0.2 m NaCl in R buffer and loaded onto a column with denatured DNA cellulose resin (Amersham Biosciences) that had been pre-equilibrated with 0.2 m NaCl in R buffer The proteins adsorbed on the DNA affinity column were eluted with 0.5 m NaCl in R buffer For more fractionation, the DNA affinity fraction was diluted to 0.1 m NaCl FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3071 Aggregation-coupled DNA end-joining M Takahagi and K Tatsumi with R buffer and was applied to a Mono Q HR ⁄ column (Amersham Biosciences) The proteins bound in R buffer including 0.1 m NaCl were eluted with 0.5 m NaCl for assays Alternatively, for purification of proteins, the Mono Q-bound proteins were developed with a 0.1–1.0 m NaCl linear gradient in R buffer This procedure generated distinct peaks, each of which was individually recovered and applied to a Mini S PC 3.2 ⁄ column (Amersham Biosciences) in R buffer containing 0.1 m NaCl (Fig S3) Each of the adsorbates, including DNA-dependent protein kinase catalytic subunit (DNA-PKcs), nucleolin, an N-terminal truncated nucleolin (nucleolinD), high mobility group box (HMGB) and HMGB2, was eluted from the column with a 0.1–1.0 m NaCl gradient in R buffer Also, the Mono S unbound proteins, including SAF-A and Ku, were recovered in the flow-through fractions Finally, these proteins were loaded onto a Mini Q PC 3.2 ⁄ column (Amersham Biosciences), and were isolated as a single peak fraction by a 0.1–1.0 m NaCl linear gradient in R buffer Histone H1, which exists in the DNA affinity fraction, was separated as a major component stepwise elution from 0.2 to 0.4 m NaCl on a tRNA-affinity column, which was prepared by coupling bovine liver-derived tRNA (Sigma, St Louis, MO, USA) to HiTrap NHS-activated HP (Amersham Biosciences) Histone H1 was finally purified by size exclusion chromatography on a Superdex 200 PC 3.2 ⁄ 30 column (Amersham Biosciences) in R buffer with 0.5 m NaCl Aliquots of fractions and isolated proteins were stored at )80 °C Before use, they were concentrated and bufferexchanged on centrifugal filter devices (Microcon YM-10, Millipore, Billerica, MA, USA), and were then separated from nonspecific aggregates by centrifugation at 8000 g The concentration of proteins was determined by the NanoOrange Protein Quantitation kit (Molecular Probes, Eugene, OR, USA) using bovine serum albumin as a standard Protein identification The N-terminal structures of proteins, including nucleolin, Ku80, nucleolinD, HMGB1 and HMGB2, were determined with an Applied Biosystems 492 model sequencer according to the instructions For proteins that might be modified in the N-terminus, including DNA-PKcs, SAF-A, Ku70 and histone H1, following treatment with sequencing grade trypsin (Promega, Madison, WI, USA), the fragmental polypeptides were isolated through reverse-phase column chromatography (lRPC C2 ⁄ C18 column, Amersham Biosciences) and were identified Osaka, Japan) Cellular fractions or isolated proteins were incubated with DNA substrates in the presence of 10 mm MgCl2 for 10 at 20 °C The mixtures were then centrifuged at 8000 g for 10 at °C The supernatants were removed or recovered, and the precipitable aggregates, after addition of the reaction buffer (100 lL), were again centrifuged After one more rinse, the aggregates were deproteinized by incubation with 0.5% SDS and 0.5 mgỈmL)1 proteinase K for 15 at 65 °C and subsequently for h at 37 °C for DNA analysis, and were analyzed by 1% agarose gel electrophoresis followed by staining with ethidium bromide For protein analysis, the aggregates were heatdenatured in loading buffer with SDS and were subjected to 10% SDS–polyacrylamide gel electrophoresis followed by staining with Coomassie brilliant blue DNA end-joining EJ reactions (20 lL) of linearized pUC18 DNA catalyzed by cellular fractions or by their resultant induced aggregates were performed in R buffer containing 10 mm MgCl2, mm ATP, 50 mm NaCl, and 0.05% digitonin at 20 °C for an hour or the indicated times After the reaction was stopped by the addition of 0.5% SDS and 0.4 mgỈmL)1 proteinase K, products were analyzed by 1% agarose gel electrophoresis in TAE buffer (40 mm Tris-acetate, pH 7.8, mm EDTA) DNA was stained with ethidium bromide Visualization and quantification of products were conducted with an EX molecular imager device (Bio-Rad, Hercules, CA, USA) Effect of a protein kinase inhibitor on EJ Mono Q bound fractions (1.4 lg) were pretreated with lm of a DNA-PK kinase inhibitor wortmannin (Sigma) for 10 at 20 °C Using the inactivated fractions, the protein–DNA aggregative complex was prepared by centrifugation as described above The aggregates were separated from supernatants and repeatedly rinsed with reaction buffer, resulting in the removal of excess inhibitor For EJ, the aggregates were incubated with mm of ATP for h at 20 °C and terminated by incubation with 0.5% SDS and 0.5 mgỈmL)1 proteinase K Additionally, after the wortmannin-treated aggregates were incubated with increasing amounts of purified DNA-PKcs for 10 at 20 °C, EJ reactions were pursued in the presence of ATP, and its complementary effect was examined DNA-PK activity Nucleoprotein–DNA coaggregation Analysis of large nucleoprotein–DNA complexes was performed basically as described previously [14,15] Reactions (20 lL) were carried out in R buffer containing 150 mm or 50 mm NaCl, and 0.05% digitonin (Wako Pure Chemical, 3072 Protein phosphorylation was examined under the conditions similar to the EJ reaction A combination of DNAPKcs (1.6 lg) and Ku70 ⁄ Ku80 heterodimer (1.0 lg) or mono Q bound fractions (1.4 lg) was used as a source of FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS M Takahagi and K Tatsumi DNA-PK kinase After centrifugal aggregation with linear DNA (0.1 lg), DNA-PK-mediated protein phosphorylarion, including its autophosphorylation, were initiated by the addition of mm ATP containing its c-32P-labeled derivative Reactions were carried out for 20 at 20 °C and stopped by incubation with SDS ⁄ PAGE loading buffer Also, the inhibitory effect of wortmannin on DNA-PK activity and its complementation by active DNA-PKcs were examined under aggregative conditions The proteins were separated on a 4–20% SDS ⁄ PAGE and the radioactive ones were visualized on an imaging plate with BAS 1800II (Fujifilm, Tokyo, Japan) Immunoblotting Proteins were resolved on a 4–12% gradient polyacrylamide gel, and were transferred onto polyvinylidene difluoride membrane The proteins reacted with first antibodies were labeled by ECL plus western blotting kit (Amersham Biosciences) and were detected on FP-3000B film (Fujifilm) Antibodies used were as follows; rabbit anti-ligase IV IgG (Oxford Biotechnology, Kidlington, UK) and rabbit antiXRCC4 antibody (Serotec, Kidlington, UK) Recombinant products of DNA ligase IV and XRCC4 (Trevigen, Gaithersburg, MD, USA) were used as positive controls Acknowledgements We thank I Furuno-Fukushi and Y Hoki for technical assistance and B S Strauss for reading the manuscript This work was supported by the Funds for Cross-over Research on Underlying Technology of Nuclear Energy and the Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology of Japan Aggregation-coupled DNA end-joining 10 11 12 13 14 15 16 References Critchlow SE & Jackson SP (1998) DNA end-joining: from yeast to man Trends Biochem Sci 23, 394–398 Peterson CL & Cote J (2004) Cellular machineries for chromosomal DNA repair Genes Dev 18, 602–616 Rogakou EP, Pilch DR, Orr AH, Ivanova VS & Bonner WM (1998) DNA double-stranded breaks induce histone H2AX phosphorylation on serine 139 J Biol Chem 273, 5858–5868 Paull TT, Rogakou EP, Yamazaki V, Kirchgessner CU, Gellert M & Bonner WM (2000) A critical role for histone H2AX in recruitment of repair factors to nuclear foci after DNA damage Curr Biol 10, 886–895 Celeste A, Fernandez-Capetillo O, Kruhlak MJ, Pilch DR, Staudt DW, Lee A, Bonner RF, Bonner WM & Nussenzweig A (2003) Histone H2AX phosphorylation 17 18 19 20 is dispensable for the initial recognition of DNA breaks Nat Cell Biol 5, 675–679 Sedelnikova OA, Pilch DR, Redon C & Bonner WM (2003) Histone H2AX in DNA damage and repair Cancer Biol Ther 2, 233–235 Aten JA, Stap J, Krawczyk PM, van Oven CH, Hoebe RA, Essers J & Kanaar R (2004) Dynamics of DNA double-strand breaks revealed by clustering of damaged chromosome domains Science 303, 92–95 Rodgers W, Jordan SJ & Capra JD (2002) Transient association of Ku with nuclear substrates characterized using fluorescence photobleaching J Immunol 168, 2348–2355 Lukas C, Falck J, Bartkova J, Bartek J & Lukas J (2003) Distinct spatiotemporal dynamics of mammalian checkpoint regulators induced by DNA damage Nat Cell Biol 5, 255–260 Bakkenist CJ & Kastan MB (2003) DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation Nature 421, 499–506 Roti Roti JL & Wright WD (1987) Visualization of DNA loops in nucleoids from HeLa cells: assays for DNA damage and repair Cytometry 8, 461–467 Malyapa RS, Wright WD & Roti Roti JL (1996) DNA supercoiling changes and nucleoid protein composition in a group of L5178Y cells of varying radiosensitivity Radiat Res 145, 239–242 Malyapa RS, Wright WD, Taylor YC & Roti Roti JL (1996) DNA supercoiling changes and nuclear matrixassociated proteins: possible role in oncogene-mediated radioresistance Int J Radiat Oncol Biol Phys 35, 963–973 Izaurralde E, Kas E & Laemmli UK (1989) Highly preferential nucleation of histone H1 assembly on scaffoldassociated regions J Mol Biol 210, 573–585 Adachi Y, Kas E & Laemmli UK (1989) Preferential, cooperative binding of DNA topoisomerase II to scaffold-associated regions EMBO J 8, 3997–4006 Luderus ME, de Graaf A, Mattia E, den Blaauwen JL, Grande MA, de Jong L & van Driel R (1992) Binding of matrix attachment regions to lamin B1 Cell 70, 949–959 Romig H, Fackelmayer FO, Renz A, Ramsperger U & Richter A (1992) Characterization of SAF-A, a novel nuclear DNA binding protein from HeLa cells with high affinity for nuclear matrix ⁄ scaffold attachment DNA elements EMBO J 11, 3431–3440 Krasnow MA & Cozzarelli NR (1982) Catenation of DNA rings by topoisomerases Mechanism of control by spermidine J Biol Chem 257, 2687–2693 Baeza I, Gariglio P, Rangel LM, Chavez P, Cervantes L, Arguello C, Wong C & Montanez C (1987) Electron microscopy and biochemical properties of polyaminecompacted DNA Biochemistry 26, 6387–6392 Tsang SS, Chow SA & Radding CM (1985) Networks of DNA and RecA protein are intermediates in homologous pairing Biochemistry 24, 3226–3232 FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3073 Aggregation-coupled DNA end-joining M Takahagi and K Tatsumi 21 Gonda DK & Radding CM (1986) The mechanism of the search for homology promoted by recA protein Facilitated diffusion within nucleoprotein networks J Biol Chem 261, 13087–13096 22 Pelta J, Livolant F & Sikorav JL (1996) DNA aggregation induced by polyamines and cobalthexamine J Biol Chem 271, 5656–5662 23 Oparin AI (1965) The origin of life and the origin of enzymes Adv Enzymol Relat Areas Mol Biol 27, 347–380 24 Huang J & Dynan WS (2002) Reconstitution of the mammalian DNA double-strand break end-joining reaction reveals a requirement for an Mre11 ⁄ Rad50 ⁄ NBS1containing fraction Nucleic Acids Res 30, 667–674 25 Audebert M, Salles B & Calsou P (2004) Involvement of poly(ADP-ribose) polymerase-1 and XRCC1 ⁄ DNA ligase III in an alternative route for DNA double-strand breaks rejoining J Biol Chem 279, 55117–55126 26 Wang H, Rosidi B, Perrault R, Wang M, Zhang L, Windhofer F & Iliakis G (2005) DNA ligase III as a candidate component of backup pathways of nonhomologous end joining Cancer Res 65, 4020–4030 27 Bloomfield VA (1996) DNA condensation Curr Opin Struct Biol 6, 334–341 28 Erard M, Lakhdar-Ghazal F & Amalric F (1990) Repeat peptide motifs which contain beta-turns and modulate DNA condensation in chromatin Eur J Biochem 191, 19–26 29 Erard MS, Belenguer P, Caizergues-Ferrer M, Pantaloni A & Amalric F (1988) A major nucleolar protein, nucleolin, induces chromatin decondensation by binding to histone H1 Eur J Biochem 175, 525–530 30 Smith GC & Jackson SP (1999) The DNA-dependent protein kinase Genes Dev 13, 916–934 31 Hammarsten O & Chu G (1998) DNA-dependent protein kinase: DNA binding and activation in the absence of Ku Proc Natl Acad Sci USA 95, 525–530 32 Hammarsten O, DeFazio LG & Chu G (2000) Activation of DNA-dependent protein kinase by singlestranded DNA ends J Biol Chem 275, 1541–1550 33 Merkle D, Douglas P, Moorhead GB, Leonenko Z, Yu Y, Cramb D, Bazett-Jones DP & Lees-Miller SP (2002) The DNA-dependent protein kinase interacts with DNA to form a protein-DNA complex that is disrupted by phosphorylation Biochemistry 41, 12706–12714 34 Sarkaria JN, Tibbetts RS, Busby EC, Kennedy AP, Hill DE & Abraham RT (1998) Inhibition of phosphoinositide 3-kinase related kinases by the radiosensitizing agent wortmannin Cancer Res 58, 4375– 4382 35 Wymann MP, Bulgarelli-Leva G, Zvelebil MJ, Pirola L, Vanhaesebroeck B, Waterfield MD & Panayotou G (1996) Wortmannin inactivates phosphoinositide 3-kinase by covalent modification of Lys-802, a residue involved in the phosphate transfer reaction Mol Cell Biol 16, 1722–1733 3074 36 Baumann P & West SC (1998) DNA end-joining catalyzed by human cell-free extracts Proc Natl Acad Sci USA 95, 14066–14070 37 Calsou P, Frit P, Humbert O, Muller C, Chen DJ & Salles B (1999) The DNA-dependent protein kinase catalytic activity regulates DNA end processing by means of Ku entry into DNA J Biol Chem 274, 7848–7856 38 Bode J, Kohwi Y, Dickinson L, Joh T, Klehr D, Mielke C & Kohwi-Shigematsu T (1992) Biological significance of unwinding capability of nuclear matrix-associating DNAs Science 255, 195–197 39 Mauldin SK, Getts RC, Liu W & Stamato TD (2002) DNA-PK-dependent binding of DNA ends to plasmids containing nuclear matrix attachment region DNA sequences: evidence for assembly of a repair complex Nucleic Acids Res 30, 4075–4087 40 Schwarz PM & Hansen JC (1994) Formation and stability of higher order chromatin structures: contributions of the histone octamer J Biol Chem 269, 16284–16289 41 Schwarz PM, Felthauser A, Fletcher TM & Hansen JC (1996) Reversible oligonucleosome self-association: dependence on divalent cations and core histone tail domains Biochemistry 35, 4009–4015 42 Guo CY, Wang Y, Brautigan DL & Larner JM (1999) Histone H1 dephosphorylation is mediated through a radiation-induced signal transduction pathway dependent on ATM J Biol Chem 274, 18715–18720 43 Kysela B, Chovanec M & Jeggo PA (2005) Phosphorylation of linker histones by DNA-dependent protein kinase is required for DNA ligase IV-dependent ligation in the presence of histone H1 Proc Natl Acad Sci USA 102, 1877–1882 44 DeFazio LG, Stansel RM, Griffith JD & Chu G (2002) Synapsis of DNA ends by DNA-dependent protein kinase EMBO J 21, 3192–3200 45 Chan DW, Chen BP, Prithivirajsingh S, Kurimasa A, Story MD, Qin J & Chen DJ (2002) Autophosphorylation of the DNA-dependent protein kinase catalytic subunit is required for rejoining of DNA double-strand breaks Genes Dev 16, 2333–2338 46 Stiff T, O’Driscoll M, Rief N, Iwabuchi K, Lobrich M & Jeggo PA (2004) ATM and DNA-PK function redundantly to phosphorylate H2AX after exposure to ionizing radiation Cancer Res 64, 2390–2396 47 Johnston PJ & Bryant PE (1994) A component of DNA double-strand break repair is dependent on the spatial orientation of the lesions within the higherorder structures of chromatin Int J Radiat Biol 66, 531–536 48 Johnston PJ, MacPhail SH, Stamato TD, Kirchgessner CU & Olive PL (1998) Higher-order chromatin structure-dependent repair of DNA double-strand breaks: involvement of the V(D)J recombination double-strand break repair pathway Radiat Res 149, 455–462 FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS M Takahagi and K Tatsumi 49 Dignam JD, Lebovitz RM & Roeder RG (1983) Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei Nucleic Acids Res 11, 1475–1489 Supplementary material The following supplementary material is available online: Fig S1 EJ process occurs within the aggregates An aggregative body consisting of the nuclear fractions and linear DNA was detached from the internal surface of a test tube by repeated pipetting The EJ activity in the free aggregates was compared with the originally fixed one (compare lanes and 5) In the course of this experiment, irrespective of the presence of the supernatant from the initial centrifugation, the yield of EJ products did not appreciably change (compare lanes and 2) The replacement of the supernatant with a reaction buffer resulted in a slight increase in the production of the multimers (compare lanes and 5) Experiments to separate the aggregate and the supernatant (sup.) after h indicated that both input substrates and the EJ products remained stable in the aggregates (compare lanes and and lanes and 6) Fig S2 The aggregation-coupled EJ activity is independent of DNA ligase IV ⁄ XRCC4 Chromatographically separated fractions were analyzed by immunoblotting using antibodies for DNA ligase IV and XRCC4 Their antigens were detectable in nuclear fraction (NF) (lane 2), but not in DNA affinity- and Mono Q-fractions that EJ activity are enriched (lanes and 4) On the other hand, the ligase components were found in a Heparin column fraction, which was Aggregation-coupled DNA end-joining separated from a flow-through fraction of the DNA affinity column (lane 5) Therefore, the EJ activity is thought to depend on a factor other than DNA ligase IV ⁄ XRCC4 complex Recombinant proteins were used to check the antibodies (lane 1) Fig S3 The aggregative EJ activity was enriched along with DNA-PKcs (A) The fractions containing major proteins were separated as distinct peaks through Mono Q column (B) Major proteins in fractions were resolved on 10% SDS ⁄ PAGE In this panel, the separation of DNA-PKcs (lane 2) and Ku70 ⁄ Ku80 (lane 3) was confirmed (C) EJ activity in Mono Q fractions was tested DNA-PKcs-enriched fraction (lane 2; marked with asterisk) contained the highest rejoining activity of pUC18-linearlized DNA (lane 2) (D) The aggregation-coupled EJ in the DNA-PKcs-enriched fraction was characterized The DNA-PKcs-fraction not only coaggregated with linear DNA, but also showed an aggregation-enhanced EJ This EJ activity was strongly inhibited by wortmannin in the presence of purified Ku70 ⁄ Ku80 proteins, implicating that DNA-PK and a ligase activity are closely interactive in aggregates It should be noted that when the DNAPKcs-enriched fraction was further fractionated through a Mini S column (Experimental procedures), a major peak fraction composed of DNA-PKcs retained a significant level of EJ activity, which was aggregative with DNA-PKcs (data not shown) The additional step using a Mini Q column allowed isolating DNA-PKcs from the concomitant EJ activity This material is available as part of the online article from http://www.blackwell-synergy.com FEBS Journal 273 (2006) 3063–3075 ª 2006 The Authors Journal compilation ª 2006 FEBS 3075 ... supports the view that the DNAPKcs -mediated aggregates stay in a catalytically competent state A further concern is whether the protein kinase activity of DNA- PKcs is really associated with the EJ... types of DNA in a Mg2+-dependent manner This preference is conditional since supercoiled DNA was aggregative in the absence of Mg2+ These results suggest that the DNA repair factor DNA- PKcs is an... SSC and SC The effect of Ku on DNA- PKcs -DNA coaggregation is indicated by a bracket (upper panel) (B) Aggregation is compatible with the kinase activity of DNA- PKcs The combination of DNA- PKcs