1

ARCHITECTURE OF THE BACTERIOPHAGE T4 REPLICATION

COMPLEX REVEALED WITH NANOSCALE BIOPOINTERSNancy G. Nossal*, Alexander M. Makhov‡, Paul D. Chastain, II‡, Charles E. Jones*, and

Jack D. Griffith‡

From the * Laboratory of Molecular and Cellular Biology, National Institute of Diabetes

and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland

20892-0830, and the ‡Lineberger Comprehensive Cancer Center, University of North

Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7295

Running Title- Architecture of the Phage T4 Replication Complex

Address correspondence to: Jack D. Griffith, Lineberger Comprehensive Cancer Center, University ofNorth Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7295, Tel. 919-966-2151, Fax 919-66-3015, E-Mail jdg@med.unc.edu.

Our previous electron microscopy ofDNA replicated by the bacteriophage T4proteins showed a single complex at thefork, thought to contain the leading andlagging strand proteins, as well as theprotein-covered single-stranded DNA on thelagging strand folded into a compactstructure. "Trombone" loops formed fromnascent lagging strand fragments werepresent on a majority of the replicatingmolecules (Chastain, P., Makhov, A. M.,Nossal, N. G., and Griffith, J. D. (2003) J.Biol. Chem. 278, 21276-21285). Here we probethe composition of this replication complexusing nanoscale DNA biopointers to showthe location of biotin-tagged replicationproteins. We find that a large fraction of themolecules with a trombone loop had twopointers to polymerase, providing strongevidence that the leading and lagging strandpolymerases are together in the replicationcomplex. Six % of the molecules had twoloops, and 31% of these had three pointers tobiotin-tagged polymerase, suggesting thatthe two loops result from two fragments thatare being extended simultaneously. Underfixation conditions that extend the laggingstrand, occasional molecules show twonascent lagging strand fragments, eachbeing elongated by a biotin-taggedpolymerase. T4 41 helicase is present in thecomplex on a large fraction of activelyreplicating molecules, but a smaller fractionof molecules with a stalled polymerase.

Unexpectedly, we found that 59 helicaseloading protein remains on the fork afterloading the helicase, and is present o nmolecules with extensive replication.

DNA replication is catalyzed bycomplexes of replication proteins that worktogether to catalyze the continuous synthesis ofthe new leading strand, and the priming,discontinuous synthesis, and joining of shortfragments on the new lagging strand. Thecoordinated interaction of these replicationproteins with each other, and with DNA, isresponsible for the accurate and timely duplicationof the genome. Because the two strands of theduplex have opposite polarity, the leading andlagging strand polymerases move in oppositedirections relative to the fork. Bruce Albertsproposed that leading and lagging strand synthesiscould be coordinated if the nascent lagging strandfragment and the single-stranded DNA behind itwere folded into what has been called a "tromboneloop" to bring the two polymerases together (Fig.1) (1).

Our previous electron microscopic studiesof the bacteriophage T4 replication complexconfirmed this trombone model (2). We found asingle complex of the leading and lagging strandproteins at the fork, with a single loop present on43% of the molecules. Another 43% of themolecules did not have a loop, as expected formolecules stopped at the stage when the previousfragment has finished, but the next fragment hadnot started. Unexpectedly, and not anticipated by

http://www.jbc.org/cgi/doi/10.1074/jbc.M606772200The latest version is at JBC Papers in Press. Published on November 13, 2006 as Manuscript M606772200

Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2

the Alberts model, 8% of the molecules had twoloops near the fork, and 5% contained more thantwo, consistent with molecules in which a newlagging strand fragment has been initiated beforethe previous fragment was completed. In contrastto the original model (Fig. 1A), the DNA in theloops was completely double-stranded, with novisible extended single-strandedDNA. Instead, theprotein-covered single-strandedDNA segments onthe lagging strand were folded into highly compactstructures (“bobbins”), which constitute the majorportion of the mass of the replication complex(Fig. 1B). This compact structure forms as a resultof lagging strand synthesis. It was not observed inthe absence of primase. Similar compact structureswere found on molecules replicated with phage T7proteins (3-5). Although the size of the complexat the fork was large enough to contain all theproteins needed for leading and lagging strandsynthesis, as well as about 800 b

1 of protein-covered single-stranded DNA, the proteinsactually present in these complexes could not bedetermined by this technique.

The T4 replication proteins provide anattractive model system for determining thearchitecture of the replication fork, and themechanisms responsible for controlling andcoordinating DNA synthesis on the two strands.This relatively simple multienzyme replication

system composed of highly purified bacteriophageT4 encoded proteins, is organized into the samefunctional enzyme groups as those in morecomplex eukaryotic replication systems (reviewedin (6,7)). T4 DNA polymerase, which catalyzesDNA synthesis on both leading and laggingstrands, is attached to a sliding clamp protein(gene 45), loaded by the complex of the gene 44and 62 proteins (Fig. 1) (8,9). Gene 41 helicasemoves 5' to 3' on the lagging strand template(10), opening the duplex ahead of the leadingstrand polymerase, and interacting with theprimase to allow it to make the RNA primers thatinitiate lagging strand synthesis (11-14). Althoughthe helicase can load on nicked and forked DNAby itself, its loading is greatly accelerated by the59 helicase loading protein (15-19). There isrecent evidence that 59 helicase loader, a forkbinding protein, plays a role in coordinatingleading and lagging strand synthesis by blockingleading strand synthesis in the absence of helicaseor 32 protein (20-24). The RNA primers andadjacent DNA are ultimately removed by a T 4

encoded 5' to 3' nuclease (T4 RNase H), and theadjacent fragments joined by T4 DNA ligase (25-27). T4 gene 32 single-stranded DNA bindingprotein coats the single-stranded DNA on thelagging strand (28), and binds and modulates theactivities of the polymerase, primase, helicaseloader and RNase H.

We have now further developeda powerfultechnique first applied to yeast Orc complexes(29) to use DNA "pointers" to biotin-taggedreplication proteins to determine which proteinsare present in the replication complexes visualizedby electron microscopy at different stages ofreplication. We show that many of the proteincomplexes at the fork contain two polymerases,as expected if the leading and lagging strandpolymerases are coordinated. Under fixationconditions that extend the lagging strand,occasional molecules show two nascent laggingstrand fragments, each being elongated by abiotin-tagged polymerase, consistent with ourprevious finding of molecules with more than onelagging strand loop (2). We find that the 59helicase loading protein remains on the fork afterloading the helicase, and is present on moleculeswith extensive replication. Biotin-tagged helicaseis present on a large fraction of activelyreplicating molecules, but on a smaller fraction ofmolecules with a stalled polymerase.

EXPERIMENTAL PROCEDURES

Cloning and expression of biotin-taggedT4 replication proteins- A sixteen residue peptide(SGLNDIFEAQKIEWHE) that is biotinylated onthe lysine in vivo by the E. coli biotin ligase(BirA) enzyme (30), followed by a gly4 or apro4gly linker, was inserted after the N-terminalmethionine of T4 gene 41 helicase (p41Nbioglyand p41Nbiopro), and T4 gene 59 helicase loadingprotein (p59Nbiogly). This was accomplished byinserting the following oligonucleotides into theNdeI site at the beginning of these genes in T 7expression plasmids.

G l y t o p : 5 'tatgtccggtctgaacgacatcttcgaagctcagaaaatcgaatggcacgaaggtggtggtggG l y b o t : 3 'acaggccagacttgctgtagaagcttcgagtcttttagcttaccgtgct

tccaccaccaccat

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

3

Protop:5'tatgtccggtctgaacgacatcttcgaagctcagaaaatcgaatggcacgaaccgccgccgccgggP r o b o t : 3 'acaggccagacttgctgtagaagcttcgagtcttttagcttaccgtgcttggcggcggcggcccat

For T4 41 helicase, an NdeI site was added topNN4101(31) by removing the small BsaXIfragment, and replacing it with oligonucleotidesthat changed the agtgtg at the beginning of gene41 to the NdeI sequence catatg. The internal NdeIsite in the gene 59 expression plasmid pNN2859(18) was removed by site directed mutagenesis.Since the N-terminus of the gene 59 helicaseloading protein is in a beta sheet in the crystalstructure (32), we also made 59 protein with itsbiotin tag at the C-terminus by addingoligonucleotides encoding the gly4 linker followedby SGLNDIFEAQKIEWHE at the end of gene 59in pNN2859 (p59Cbiogly).

To introduce the same sequences at theN-terminus of T4 DNA polymerase, a Bsg I sitewas first introduced just after the Bam HI site inpPST4pol (33), by amplifying the regionbetween the Bam HI and Xho I sites using theprimers 5'GCAGGATCCGTGCAGACTAAGGAATATCTATG (43 PCR TOP) and 5'CGCTTCATCCAATCTCGAGCATCTTTCATTG (43 PCR BOT). The Bam HI to Xho Iregion of the PCR fragment was then used toreplace the same region in pPST4pol(pRB405). Finally the Bsg I to Pst I fragment ofpRB405, the Pst to Nde I fragment of the T7expression vector pVex11, and oligonucleotidesencoding SGLNDIFEAQKIEWHE followed by agly4 or a pro4gly linker were ligated together(p43Nbiogly and p43Nbiopro).

For expression, we first isolated pBirAcm,a pACYC184 plasmid with an IPTG inducible birAgene to overexpress biotin-protein ligase, from E.c o l i AVB99, purchased from Avidity(www.avidity.com). The pBirAcm plasmid wasthen transformed into E. coli BL21 (DE3), withchloramphenicol at 10 μg/mL. Finally we moved

each of the plasmids encoding T4 biotin fusionproteins into E. coli BL21 (DE3) (pBirAcm).Cultures were grown to an OD600 of 0.4 at 37

o Cin LB media supplemented with 100 μM final

concentration biotin, 10 μg/ml chloramphenicol,

and 50 μg/ml carbenicillin. Addition of IPTG t o

a final concentration of 1 mM induces bothbiotin-protein ligase and T7 RNA polymerase, andgave good expression of each of the T4 fusionproteins after two hours.

Purification of biotin-tagged T4replication proteins- Proteins were purified bymodifications of the methods previously describedfor the unmodified proteins, followed by affinitychromatography on monovalent avidin (Softlinkfrom Promega) for the helicase and polymerase,or the higher capacity Streptavidin Mutein Matrix(Roche) for 59 protein. Biotin in the proteins wasmeasured using streptavidin horseradish peroxidaseand the Protein Detector LumiGLO Western BlotKit from KPL, Inc.

The N- and C- terminally tagged biotin-59 proteins with gly4 linkers from 2 liter cultureswere partially purified through the high saltsupernatant step (34). The affinity purificationon the Streptavidin Mutein Matrix, by thefollowing modification of the manufacturer'sprotocol, was carried out at 4°C. A 4 ml column(stated capacity is 2.5 mg biotinylated protein perml), was washed with 40 ml of wash buffer (100mM potassium phosphate, pH 7.2, 150 mMNaCl), and then with 16 ml of equilibration buffer(100 mM potassium phosphate, pH 7.2, 150 mMNaCl, and 400 mM ammonium sulfate). Half ofthe high salt supernatant (2 ml) was mixed with1.25 ml 3X equilibration buffer and then appliedto this column. The column was closed for 40minutes to allow protein to bind, and then washedwith 40 ml of the wash buffer. Four ml of elutionbuffer (wash buffer with 2 mM D-biotin) wereallowed to run into the column, the column closedfor 30 minutes, and the biotin-tagged protein theneluted with 20 ml more of the elution buffer,collecting 0.5 ml fractions. Most of the biotintagged 59 protein was in the first 2 ml. Thematrix was regenerated following themanufacturer's protocol, and then used to purifythe remainder of the 59 protein. Peak fractionswere pooled, dialyzed against 50 mM Tris-HCl,pH 7.5, 10% glycerol, 1 mM EDTA, 0.5 mMTCEP, and 100 mM KCl, and stored at -85

o C.The N-terminal biotin-tagged T4 DNA

polymerases from p43Nbiogly and p43Nbioprowere purified from 2 l cultures by the rapid batchphosphocellulose method described previously(35), and then on a Softlink monovalent biotinresin from Promega. To saturate nonspecificbinding sites for biotin and regenerate the resin, 1

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

4

ml columns were washed sequentially with 5 ml0.1 M NaPO4 7.0; 5 ml of the same buffercontaining 10 mM biotin; 10 ml 10 % acetic acid;and then with 10 ml 0.1 M NaPO4 7.0 to bringthe pH of the effluent to 6.8. The column wasclosed for 1 hour at room temperature to allowthe avidin to refold, moved to 4 °C, andequilibrated with 43 SL buffer (50 mM Tris-Cl, pH7.5, 1 mM EDTA, 10% glycerol, 0.1M KCl and 1mM DTT). Polymerase (2 mg) was loaded in 0.5ml aliquots, closing the column for 15 min aftereach addition. The column was then washed with8 ml 43SL buffer, and the biotin-tagged proteineluted with 43SL buffer containing 10 mM biotin,closing the column for 15 min after each 0.5 mladdition. Most of the protein (40 % recovery) wasin the second and third 0.5 ml fractions.Polymerase was dialyzed against 43 SL buffer andstored at -85°C.

The N-terminal biotin-tagged T4 41helicase from a 5 liter culture of p41biopro orp41biogly was purified through the Q-Sepharosechromatography step as described (31). A 1 mlcolumn of Softlink monovalent avidin was treatedas described above, and then equilibrated with ATbuffer (10% glycerol, 50 mM NaTAPS, pH 8.5,0.5 mM TCEP-HCl (Pierce)). A peak fraction ofhelicase from the Q-Sepharose column (about 1.5mg) was loaded on this Softlink column, thecolumn washed with 1 ml AT buffer, biotin-helicase eluted with AT buffer containing 5 mMbiotin, dialyzed against AT buffer, and stored at -85°C.

Nicked DNA templates- Plasmid pUCNICK(2716 bp) with a single recognition site for theN.BbvC IA nicking enzyme (New EnglandBiolabs) was constructed, purified and nicked asdescribed previously (34). pNNBSGless, which hasa recognition sequence for the N.BbvC1A nickingenzyme followed by a 396 bp cassette with no Cin the top strand, has the sequenceGAATTTTAAGTAGGTTAAGGGGTTAAGC|TG

AGG (N.BbvC1A recognition sequence underlined,

nicked site indicated by |) inserted between the RIand Sma sites of pBSGLess (36). Theconstruction of the 452 bp minicircle, whichcontains a recognition site for the NBstN1Bnicking enzyme (New England Biolabs) followedby six copies of a 70 bp sequence with the all theG on one strand, will be described elsewhere.

Replication reactions, fixation, andaddition of the biotin pointers- T4 DNA ligase was

obtained from USB Biochemicals. Thepurification of T4 32 protein (20), and all otherunmodified T4 replication proteins (35) has beendescribed previously. Replication reactionmixtures (40 μl) contained 2 nM of the nicked

DNA templates, 2mM ATP, 250 μM of each

dNTP, 250 μM CTP, GTP, and UTP, 25 mM K

Hepes (pH 7.6), 60 mM potassium acetate, 6 mMmagnesium acetate, 10 mM �-mercaptoethanol,

and 20 μg/ml bovine serum albumin. Enzymes

were diluted in a solution containing 50 mM KHepes (pH 7.6), 100 mM KCl, 5 mM MgCl2,1mMEDTA,10mM �-mercaptoethanol, 100 μg/ml

bovine serum albumin, and 25% glycerol. Unlessotherwise noted, the protein concentrations were2 μM gene 32 ssDNA binding protein, 328 nM

gene 41 helicase, 30 nM DNA polymerase, 242nM genes 44/62 clamp-loader, 162 nM gene 45clamp, 100 nM gene 59 helicase loading protein,64 nM gene 61 primase, 195 nM RNase H, andDNA ligase at 75 Weiss units/ml. Reactionmixtures without proteins were incubated for 2min at 37

o C, and the reaction begun by additionof a mixture of all the proteins except T4RNaseH and ligase, which were added 30 secondslater to prevent ligation of the nicked templatesbefore replication began. Biotin-tagged replicationproteins replaced wild type proteins as indicated.Unless otherwise noted, reactions were stopped byadding 20 μ l of 45 mM EDTA, and 1.8%

glutaraldehyde in 1X replication salts withoutMgCl, giving a final concentration of 0.6 %glutaraldehyde. After 5 min at room temperature,the glutaraldehyde fixation was quenched byadding 20 μl of 400 mM Tris-Cl, pH 7.5 and 20

mM EDTA, followed 10 minutes later by 20 μl of

2.4 μM streptavidin (Molecular Probes) in 10 mM

Tris-Cl, pH 7.5, 100 mM NaCl, and 1 mM EDTAto bind the biotin-tagged protein. The sampleswere placed on a rotator for 30 min at roomtemperature, and then filtered on 2 ml columns of50-150 micron 2% agarose beads (Agarose Bead

Technologies) in 10 mM Tris-Cl, pH 7.5 and 0.1mM EDTA to remove unbound streptavidin andreplication proteins. A 96 μ l aliquot of the

fraction containing the DNA-protein complexeswas mixed with a 5' biotinylated 300 or 179 bpDNA, amplified from Bluescript plasmid, to givea final concentration of 2 μg/ml (9.6 μM). The

179 bp DNA was used in Fig. 3F. The 300 bpDNA, which is more clearly visible on the

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

5

micrographs, was used in all other figures. Thesamples were mixed on a rotator for 18-20 hoursat 4o C to allow binding of the biotinylated DNA,and then filtered on 2 ml columns of the 2%agarose in 10 mM Tris-Cl, pH 7.5 and 0.1 mMEDTA to remove unbound biotinylated "DNA

pointers".Replication reactions for analysis on gels

were carried out as described above for EM, exceptthat RNase H and ligase were omitted so that thelagging strand fragments were not joined together.At the times indicated in the figures, aliquots ofthe reaction mixtures were mixed with an equalvolume of 0.2 M EDTA to stop the synthesis, andthe products were analyzed by 0.6% alkalineagarose gel electrophoresis and trichloroaceticacid precipitation (35).

Electron microscopy - Samples wereadsorbed to thin carbon foils, washed, air-dried,and rotary shadowcast with tungsten at highvacuum (37). Samples were examined in an FEITecnai 12 instrument at 40 kV. Lengthmeasurements were made by capturing the imageswith a Gatan 4K CCD camera attached to theTecnai 12 and using Digital Micrograph software(Gatan Inc., Pleasanton CA). Images forpublication were captured on the Gatan CCD or onsheet film and then scanned with an Imacon 848film scanner, and the contrast optimized andpanels arranged using Adobe Photoshop software.We only analyzed molecules with a circle thesame size as the starting template.

RESULTS

Molecular pointers to proteins within theT4 replication complex- We have usedbiotinylated DNA as a molecular pointer t oidentify proteins within the T4 replicationcomplex. A sixteen residue peptide(SGLNDIFEAQKIEWHE) that is biotinylated onthe lysine in vivo by the E. coli biotin ligase(BirA) enzyme (30), followed by a gly4 or apro4gly linker, was inserted after the N-terminalmethionine in T4 DNA polymerase, 41 helicase,and 59 helicase loading protein, as described inexperimental procedures. The purificationmethods for each protein included an affinitycolumn (Softlink (Promega) or the highercapacity Streptavidin Mutein Matrix (Roche)) toremove proteins without the biotin tag. Thedegree of biotinylation of the helicase and

polymerase was higher with the pro4gly than thegly4 linker, suggesting that the more rigid linkermade the peptide more accessible to the biotinligase.

The biotin-tagged polymerase and 59helicase loader had activity equivalent to the wildtype proteins (Fig. 2A and C). The template is the2.7-kb pUCNICK plasmid, nicked at the singlerecognition site for the N-BbvC IA nickingenzyme (34). Reactions contained T4 DNApolymerase, 45 clamp, 44/62 clamp loader,primase, 41 helicase, 59 helicase loader, and 32single-stranded DNA binding protein. The laggingstrand fragments were not joined because no T 4RNase H or DNA ligase was added to thesereactions. 59 helicase loader with the same tag atits C-terminus retained about 50% of the wild typeactivity (data not shown). T4 41 helicase isrequired for both increasing the rate of leadingstrand synthesis by unwinding the duplex ahead ofthe polymerase, and allowing the primase to makethe pentamer primers that initiate the laggingstrand fragments. The biotin-tagged helicaseretains both these activities, and is nearly asactive as the wild type helicase (Fig. 2B).

We needed a DNA pointer that was longenough to extend beyond the large T4 replicationcomplex (2) and be clearly visible on EM images.In our early experiments we used a pointercomposed of a purified complex of a 179 bpbiotinylated DNA duplex bound to a streptavidintetramer (29). However, we have found that it iseasier and just as efficient to add streptavidin t othe fixed replicating molecules, remove unboundreplication proteins and streptavidin by gelfiltration, and then add a 179 or 300 bpbiotinylated DNA (see experimental procedures).In the replication reactions for EM, the laggingstrand fragments were joined by T4 RNaseH andDNA ligase. These were added 30 seconds afterthe other enzymes to prevent ligation of thenicked templates before replication began.

T4 replication complexes contain bothleading and lagging strand polymerases- Ourprevious EM analysis of T4 replicating moleculesshowed a single complex at the replication forkthat was large enough to contain the clampedleading and lagging strand polymerases, thehelicase, primase, and helicase loader, and about800 b of single-stranded DNA covered with 32protein folded into a compact structure (2). Therewas frequentlya double-stranded"trombone" loop,

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

6

formed from the nascent lagging strand fragment,attached to the complex. The 300 bp biotinylatedpointer clearly shows the presence of twopolymerases in this complex on the 2.7 kbcircular pUCNICK template (Fig. 3 A, B and D),and on a 452 bp nicked mini circle (seeexperimental procedures) (Fig. 3C). The minicircle was designed to be large enough to be visibleon the micrographs. In a series of six experimentswith the biotin-tagged polymerase, 27-70% of themolecules scored had one or more pointers on thecomplex, and 10-43% had two or more pointers.Control reactions with all wild type proteins hada single pointer on 3-7% of the complexes,consistent with a low level of non-specificbinding.

Although streptavidin is multivalent, thevery strong correlation between the number ofloops on the replicating molecules and the numberof pointers to biotin tagged polymerase, makes ithighly unlikely that many of the molecules hadtwo pointers to the same streptavidin. As shownin Fig. 3G, 29% of the molecules with one loop,where the leading and lagging strand polymerasesshould be together in the complex, had twopointers to polymerase. In contrast, only 3% ofthe molecules without a loop had two pointers.Molecules with two loops, which we interpret ashaving two lagging strand fragments elongatingsimultaneously, were rare (13/221 moleculesscored). Thirty-one % of these had three pointersto polymerase, while there were no molecules thathad one loop, or no loop, that had three pointers.We frequently saw molecules where the complexhad spilt into two parts, with a pointer t opolymerase on each part (Fig. 3 D and E). Thecomplex at the junction between the circulartemplate and the tailed product likely contains theleading strand proteins. The distal complex isseparated from the fork by duplex DNA, which isthe nascent lagging strand fragment. There werealso rare molecules (Fig. 3F) where the proteincovered single stranded DNA on the lagging strandwas extended, and there were pointers t opolymerases on two adjacent lagging strandf r a g m e n t s , at the expected location forpolymerase molecules completing an Okazakifragment.

Helicase in the replication complex- T 441 helicase is a hexamer of identical subunits. Inexperiments with the biotin-tagged helicase, weoccasionally saw six pointers to the hexameric

helicase (Fig. 4A), but most molecules had fewerpointers (Fig. 4 B-D). In the experiment analyzedin Fig. 4E , 91 of the 108 molecules from thereaction with the pUCNICK template had a doublestranded tail indicating there had been both leadingand lagging strand synthesis. There were 1 ormore pointers to the helicase on 55 (60%) of themolecules with double-stranded tails. The pointerswere at the fork on 52 of these molecules, withremaining 3 on the tail. In a series of threeexperiments with the tagged helicase, 60-70% ofthe molecules with double-stranded tails had atleast one pointer, 24-28% had two or morepointers, including 1-3% with six pointers. Twofactors may contribute to the small proportion ofmolecules with six pointers. It is likely that mostof the tagged-helicase had some subunits thatlacked biotin, because not all the helicase with thepeptide tag is actually biotinylated in vivo (datanot shown), and only one biotinylated subunit wasnecessary for the helicase to be bound to theaffinity column used in the purification of theprotein. In the absence of ATP, the helicase is adimer in solution, so that a minimum of threebiotinylated subunits would be expected in theactive hexamer (38). Thus our finding that only11% of the molecules had three or more pointerssuggests that the location of the helicase withinthe replication complex, or the addedstreptavidin, shields some of the subunits from thepointers. Conversely, on molecules where twopointers end at the same place in complex, theymay both be pointing to same subunit, sincestreptavidin is multivalent.

Less helicase is present on stalledreplicating molecules- The processivity of T4 41helicase is greatly increased when it is present witha rapidly moving polymerase (reviewed in(39,40)). We used a nicked plasmid(pNNGless) onwhich leading strand synthesis stalls after 396 b inthe absence of dCTP to compare biotin-taggedhelicase and polymerase on stalled and rapidlyreplicating molecules. With all 4 dNTPs, there israpid synthesis and pointers show that bothbiotin-helicase (Fig. 5A) and polymerase (Fig. 5B)are present. Multiple tagging approaches would berequired to formally demonstrate that both arepresent on the same complex. In the absence ofdCTP, the stalled replicating molecules have a396 b single-stranded tail because there are no T4primase recognition sequences on the displacedstrand. The pointers showed that both helicase

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7

and polymerase were present in these stalledcomplexes. Multiple tagging approaches would berequired to formally demonstrate that both arepresent on the same complex." However, analysisof a large number of molecules from each reaction(Fig. 5C) revealed that helicase was present ononly 24% of the stalled molecules, compared with53% of the molecules replicated with all 4 dNTPs.There was less of a difference with the taggedpolymerase, which was present on 49% of thestalled molecules and 65% of the replicatingmolecules. In the future it will be important t oobtain direct biochemical confirmation of thisobservation as it is a central issue in understandingthe nature of the stalled replication fork.

59 Helicase loading protein remains onthe DNA after extensive replication- T4 gene 59helicase loading protein accelerates the loading ofthe gene 41 helicase at the replication fork. Ourexperiments with pointers to biotin-tagged 59loader show that it remains on the fork onmolecules where there has been extensivereplication (Fig. 6). In the experiment analyzedin Fig. 6G , 96 molecules had a double-strandedtail, the product of leading and lagging strandsynthesis, and of these 50 (52%) had one or morepointers to N-terminally tagged biotin 59 loaderat the fork. Thus the percentage of replicationcomplexes with a pointer to biotin-tagged 59loader was similar to that we observed with thebiotin-tagged polymerase and helicase (Figs. 3-5).Forty-three percent of these molecules had a

single pointer to the loading protein (Fig. 6A), 8%had two pointers (Fig. 6B), and only 1% had threeor more pointers (Fig. 6C). For comparison, in aparallel experiment 30% of the molecules with adouble-stranded tail had a single pointer to biotin-tagged helicase, 16 % had two pointers, and 11%had three or more pointers (Fig. 4E).

Under our conditions, most of the protein-covered single-stranded DNA that has beenunwound by the helicase on the lagging strand atthe fork is folded into a compact structure, asshown in Fig. 6 A-C (see also (2)). There wereoccasional molecules in which this single-strandedDNA was open (Fig. 6 D-E). Pointers on thesemolecules showed that 59 helicase loader waslocated at or near the fork, as well as near thedistal end of this single-stranded region, close t othe point where the last lagging strand fragmenthad been initiated. There were also moleculeswhere 59 loader was on a protein-covered single

stranded gap separated from the fork by duplexDNA (Fig. 6F).

The number of molecules with multiplepointers to 59 helicase loader at the forkincreased with increasing 59 loader concentration,but this did not correlate with increased DNAsynthesis. Molecules with 2 or more pointerswere 4, 8-9, and 32% of the total at 50, 100, and200 nM 59 loader respectively, while moleculeswith a single pointer were 36, 39-42, and 32% inthe same reactions. However, DNA synthesis washighest at 100 nM (Fig. 2C ), consistent withprevious reports of inhibition with higher 59loader concentrations (34).

Since 59 loader is known to bind to single-stranded DNA, we carried out control reactions tomeasure its binding to 7200 b single-stranded M13DNA under our replication conditions (Fig. 7). Wewanted to be sure that 59 protein with pointers inthe replication reactions did not result from 59protein simply binding to single-stranded regionson the lagging strand template. These controlreactions had all of the proteins present in ourreplication reactions except primase, with 59protein at 100 nM. The protein-covered circularmolecules had an open "beads on a string"appearance, similar to that we have observed withthe same DNA and only 32 and 59 proteins underthe same conditions (not shown), rather than thecompact single-stranded DNA characteristic ofthe replicating molecules. Most of the visibleprotein is 32 protein. Only 36% of the moleculeshad a pointer to 59 protein, with most molecules(28%) having a single pointer. Since the availablesingle-stranded DNA on the M13 molecules isabout seven times that of the 1 kb on thereplicating molecules (2), the density of 59protein is much higher at or near the fork duringreplication (Compare Figs. 6 and 7).

DISCUSSION

We have used electron microscopycombined with biotinylated linear DNA as amolecular pointer to show that biotin tagged T 4DNA polymerase, gene 41 helicase, and gene 59helicase loading protein remain in the protein-DNA complex at the replication fork during DNAsynthesis in vitro. The presence of a pointer to aspecific biotinylated protein provides unambiguousevidence that the protein is there. However, thenumber of subunits of each protein in the complex

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

8

cannot be determined because the attachment ofthe pointers is not 100% efficient, andstreptavidin has four biotin-binding sites. The T4replication complex at the fork contains protein-coated single-stranded DNA folded into a compactstructure, in addition to the polymeraseholoenzyme and primosome proteins directlyresponsible for new DNA synthesis. We have notbeen able to use the pointers to determine therelative positions of specific tagged proteinswithin this highly dynamic complex, whose sizeand shape changes at different stages of thelagging strand cycle. In contrast, the pointers wereused successfully to show the location of proteinswithin the static complex of yeast Orc proteins,bound at a replication origin (29).

Both leading and lagging strandpolymerase are present in the replicationcomplex- We previously established that a largefraction of the T4 replication complexescontained a double-stranded loop, and that thedistribution of lengths of these loops matched thesize distribution of lagging strand fragments ondeproteinized molecules (2). This stronglysuggested that the loops were formed from thenascent lagging strand fragments and shouldcontain the lagging strand polymerase at theelongating end of the loop, in addition to theleading strand polymerase that should be presentat the junction of the circular template and thelinear rolling circle product. The studies presentedhere provide the strongest evidence that thereplication complexes with a loop do contain twopolymerases. The two polymerases are within atight complex on most of the looped molecules.However, there were several molecules in whichthe complex had opened showing two separatedcomplexes, each with a pointer to polymerase,one at the fork junction, and the other separatedfrom the fork by a duplex with a nascent laggingstrand fragment, consistent with their assignmentas the leading and lagging strand polymerases.

Molecules without a loop are at a stage inthe lagging strand cycle where a lagging strandfragment has been completed, but polymerase hasnot begun the extension of the next primer, orthe extended chain is too short to form a visibleloop. We found that 44% of the molecules witha long duplex product tail, but no loop, had asingle pointer to polymerase, and only threepercent had two pointers. This was surprisingbecause it has been shown that T4 lagging strand

synthesis can continue after dilution into asolution without polymerase (41), implying thatthe lagging strand polymerase can be recycled t othe next fragment. It is possible that when apolymerase is in transit to the next primer, or inthe initial stage of primer extension, it is moreeasily lost from the complex under our fixationconditions. The association of both the leadingand lagging strand polymerases with thereplicating DNA is dynamic, in the sense thatthey can be replaced by a mutant polymerasepresent in the reaction solution (42). It hasrecently been proposed that the presence of aclamp on a primer serves as a signal for thelagging strand polymerase to leave an unfinishedfragment to begin elongation of the clampedprimer (43). It is clear that the size of laggingstrand fragments increases with limitingconcentrations of the clamp

2 (43), as expected ifa primer must be clamped before it can beextended. If polymerase normally leaves anunfinished fragment when the clamp is notlimiting, the binding of a second polymerase t ocomplete the fragment must be very efficient.Very few gaps are observed on replicatingmolecules made under these conditions in thepresence of T4 DNA ligase and the T4 5' nuclease(RNaseH) that removes the primers ((2) and thispaper), and almost all the adjacent fragments areseparated only by nicks that can be sealed by T4DNA ligase

2.In our original analysis of molecules

replicated by the T4 proteins, we unexpectedlyfound a small percentage of molecules thatappeared to have more than one loop (2). Whenthe same reaction products were examined afterremoval of the T4 proteins by proteolysis, therewas a similar percentage of molecules with twoduplex regions surrounded by single-stranded DNAnear the fork, consistent with molecules on whichthere were two incomplete lagging strandfragments. These molecules with two loops, ortwo lagging strand fragments, were not predictedby the original Albert's trombone replicationmodel. In the present study, 6% of the moleculeshad two loops, and 31% of these had threepointers to biotin-tagged polymerase, providingstrong evidence that the two loops in fact resultfrom two fragments that are being extendedsimultaneously. There were no molecules with asingle loop or no loop that had three pointers t opolymerase.

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

9

More helicase is present on rapidlyreplicating molecules- We found one or morepointers to biotin-tagged gene 41 helicase on 60%of the molecules with double-stranded tails on thecircular templates, and almost all of these were atthe fork, rather than on unwound single-strandedDNA ahead of the polymerase. Molecules with sixpointers to this hexameric helicase of identicalsubunits were rare for reasons discussed in"Results". The T4 gene 41 and T7 gene 4replicative helicases each have been shown t ounwind at a much higher rate at a replication forkahead of the leading strand polymerase, than onmodel fork helicase substrates (reviewed in (44)).The important factor appears to be having duplexDNA behind the helicase to keep it in position,rather than a specific interaction with thepolymerase. The processive T7 polymerase-thioredoxin complex can replace T4 DNApolymerase with the T4 helicase (39,41), and T4DNA polymerase can replace the T7 DNApolymerase complex with T7 helicase (45). Wefound significantly less helicase on moleculeswhere replication was stalled after 396 b by theabsence of a required dNTP, than on rapidlyreplicating molecules in a complete reaction withthe same template.

T4 59 helicase loading protein remains onthe DNA after loading the helicase- 59 helicaseloader is present on the majority (70%) of themolecules with a long double-stranded tailreplicated in reactions with biotin-tagged 59helicase loader. Most of the pointers (52%) wereat the fork, and the 18% scored on the tail werebound predominantly to the single-stranded DNAadjacent to the fork. Since 59 protein bindssingle-stranded DNA, it was important to establishthat the binding we observed on the replicationproducts was greater than that for 59 protein onany single-stranded DNA under the sameconditions. With M13 circular single-strandedDNA at the same concentration as the nickedtemplates, and primase omitted to preventsynthesis, we found pointers to 59 protein ononly 36% of the molecules, and only 8% hadmore than one pointer. Since the single-strandedregion on the replicating molecules averages 1000bases, rather than the 7200 bases on each M13circle, the density of 59 protein bound to single-stranded DNA is clearly much greater duringreplication. Thus 59 protein remains on the DNAduring replication, after loading the helicase.

The function of 59 protein beyondincreasing the rate of helicase loading is a subjectof active investigation. This small (26 kda)protein binds preferentially to fork DNA, andinteracts directly with T4 41 helicase, 32 proteinand polymerase (17,20,23,24,46). T4 phagebegins replication from one or more replicationorigins, but most of its replication is accomplishedat forks established by recombination. Earlystudies showed that T4 phage mutants in gene 59had a DNA arrest phenotype, suggesting a role inrecombination-directed replication (reviewed in(47)). At the second fork established forbidirectional replication in vivo at a T 4replication origin, leading strand synthesiscontinued in the absence of lagging strandsynthesis with a gene 59 mutant, while synthesison the two strands began simultaneously with thewild type phage. This led to the hypothesis that59 protein functions as a gatekeeper, blocking theleading strand polymerase until the primase-helicase is loaded (22). In vitro, the presence of 59protein has been shown to block leading strandsynthesis until the helicase and 32 protein, whichare each necessary for lagging strand synthesis, areloaded (20,21,23,24). Cross linking andfluorescence transfer (FRET) experiments showa close interaction of 59 helicase loader andpolymerase (23,24). Single molecule FRETexperiments on a short primed fork (longeststrand 62 b) indicated that 59 and 32 proteinremained on the fork with helicase in the presenceof ATP�S, but both 59 and 32 protein left when

ATP, required for the assembly of the helicasehexamer and DNA unwinding, was added (19).Similar results with the helicase and 59 proteinwere obtained with gel mobility shift experiments,using antibody to identify complexes containingthe helicase loader (31). These studies all showthat 59 protein and active helicase are not presentsimultaneously on the short fork DNA. It isunclear whether there is simply not room on theshort DNA, or if 59 protein is actively ejected aspart of the helicase loading mechanism. If it isejected, it must bind efficiently to the single-stranded DNA produced as helicase unwinds theduplex at the replication fork, because we foundpointers to 59 protein at or near the fork on amajority of the replicating molecules. The role ofthis bound 59 protein is still unclear. Both 59protein and 32 protein increase the rate ofsynthesis of pentamer primers by the primase-

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

10

helicase (14). The 32 protein-covered single-stranded DNA on molecules replicated with theT4 proteins appeared less compact in reactionswithout 59 protein (2). The possibility that the

59 and 32 proteins have roles in regulating primersynthesis or utilization on the lagging strand needsto be examined.

Acknowledgements- We thank Erin Green for cloning and purifying the C-terminally tagged 59helicase loader. This work was supported by the Intramural Research Program of NIDDK, NationalInstitutes of Health (NGN) and NIH grant GM31819 (JDG).The co-authors all wish to dedicate this work to our first author, Dr. Nancy Nossal, who sadly passedaway following the completion of this work. In the finest tradition of first authorship she carried outthe greatest share of the experimental work in this paper including crafting the final product.

FOOTNOTES1 The abbreviations used are: b, base; bp, base pair;ssDNA, single-stranded DNA; dsDNA, double-stranded DNA; TAPS, [tris(hydroxymethyl)methyl]aminopropanesulfonic acid); TCEP, tris(2-carboxyethyl)-phosphine hydrochloride; DTT, dithiothreitol.2 Unpublished experiments, NGN.

REFERENCES

1. Alberts, B. M. (1985) Trends in Genetics 1, 26-30

2. Chastain, P., Makhov, A. M., Nossal, N. G., and Griffith, J. D. (2003) J Biol Chem. 278,

21276-21285

3. Lee, J., Chastain, I., Paul D., Griffith, J. D., and Richardson, C. C. (2002) J Mol Biol 316,

19-34

4. Lee, J., Chastain, P. D., 2nd, Kusakabe, T., Griffith, J. D., and Richardson, C. C. (1998)

Mol Cell 1, 1001-1010

5. Park, K., Debyser, Z., Tabor, S., Richardson, C. C., and Griffith, J. D. (1998) J Biol

Chem 273, 5260-5270

6. Nossal, N. G. (1994) in Molecular Biology of Bacteriophage T4 (Karem, J., ed), pp. 43-

53, American Society for Microbiology, Washington, D. C.

7. Benkovic, S. J., Valentine, A. M., and Salinas, F. (2001) Annu Rev Biochem 70, 181-208.

8. Alley, S. C., Trakselis, M. A., Mayer, M. U., Ishmael, F. T., Jones, A. D., and Benkovic,

S. J. (2001) J. Biol. Chem. 276, 39340-39349

9. Pietroni, P., Young, M. C., Latham, G. J., and von Hippel, P. H. (2001) J Mol Biol 309,

869-891.

10. Venkatesan, M., Silver, L. L., and Nossal, N. G. (1982) J. Biol. Chem. 257, 12426-12434

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

11

11. Cha, T.-A., and Alberts, B. M. (1990) Biochemistry 29, 1791-1798

12. Hinton, D. M., and Nossal, N. G. (1987) J. Biol. Chem. 262, 10873-10878

13. Jing, D. H., Dong, F., Latham, G. J., and von Hippel, P. H. (1999) J Biol Chem 274,

27287-27298.

14. Valentine, A. M., Ishmael, F. T., Shier, V. K., and Benkovic, S. J. (2001) Biochemistry 40,

15074-15085.

15. Barry, J., and Alberts, B. M. (1994) J. Biol. Chem. 9, 33049-33062

16. Delagoutte, E., and von Hippel, P. H. (2005) J Mol Biol 347, 257-275

17. Jones, C. E., Mueser, T. C., Dudas, K. C., Kreuzer, K. N., and Nossal, N. G. (2001) Proc

Natl Acad Sci U S A 98, 8312-8318.

18. Spacciapoli, P., and Nossal, N. G. (1994) J. Biol. Chem. 269, 447-455

19. Zhang, Z., Spiering, M. M., Trakselis, M. A., Ishmael, F. T., Xi, J., Benkovic, S. J., and

Hammes, G. G. (2005) Proc Natl Acad Sci U S A 102, 3254-3259

20. Jones, C. E., Mueser, T. C., and Nossal, N. G. (2004) J Biol Chem 279, 12067-12075

21. Nossal, N. G., Dudas, K. C., and Kreuzer, K. N. (2001) Mol Cell 7, 31-41

22. Dudas, K. C., and Kreuzer, K. N. (2005) J. Biol. Chem. 280, 21561-21569

23. Xi, J., Zhang, Z., Zhuang, Z., Yang, J., Spiering, M. M., Hammes, G. G., and Benkovic,

S. J. (2005) Biochemistry 44, 7747-7756

24. Xi, J., Zhuang, Z., Zhang, Z., Selzer, T., Spiering, M. M., Hammes, G. G., and Benkovic,

S. J. (2005) Biochemistry 44, 12264

25. Bhagwat, M., and Nossal, N. G. (2001) J Biol Chem 276, 28516-28524.

26. Gangisetty, O., Jones, C. E., Bhagwat, M., and Nossal, N. G. (2005) J. Biol. Chem. 280,

12876-12887

27. Hollingsworth, H. C., and Nossal, N. G. (1991) J. Biol. Chem. 266, 1888-1897

28. Alberts, B. M., and Frey, L. (1970) Nature 227, 1313-1318

29. Chastain, P. D., II, Bowers, J. L., Lee, D. G., Bell, S. P., and Griffith, J. D. (2004) J. Biol.

Chem. 279, 36354-36362

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

12

30. Schatz, P. J. (1993) Biotechnology (N Y) 11, 1138-1143

31. Jones, C. E., Mueser, T. C., and Nossal, N. G. (2000) J. Biol. Chem. 275, 27145-27154

32. Mueser, T. C., Jones, C. E., Nossal, N. G., and Hyde, C. C. (2000) J Mol Biol 296, 597-

612

33. Spacciapoli, P., and Nossal, N. G. (1994) J. Biol. Chem. 269, 438-446

34. Jones, C. E., Green, E. M., Stephens, J. A., Mueser, T. C., and Nossal, N. G. (2004) J

Biol Chem 279, 25721-25728

35. Nossal, N. G., Hinton, D. M., Hobbs, L. J., and Spacciapoli, P. (1995) Methods in

Enzymology 262, 560-584

36. Subramanian, D., and Griffith, J. D. (2005) J. Biol. Chem. 280, 42568-42572

37. Griffith, J. D., and Christiansen, G. (1978) Annual Review of Biophysics and

Bioengineering 7, 19-35

38. Dong, F., Gogol, E. P., and von Hippel, P. H. (1995) J. Biol. Chem. 270, 7462-7473

39. Delagoutte, E., and von Hippel, P. H. (2001) Biochemistry 40, 4459-4477.

40. von Hippel, P. H., and Delagoutte, E. (2001) Cell 104, 177-190.

41. Kadyrov, F. A., and Drake, J. W. (2001) J Biol Chem 276, 29559-29566.

42. Yang, J., Zhuang, Z., Roccasecca, R. M., Trakselis, M. A., and Benkovic, S. J. (2004)

Proc Natl Acad Sci U S A 101, 8289-8294

43. Yang, J., Nelson, S. W., and Benkovic, S. J. (2006) Mol Cell 21, 153-164

44. Patel, S. S., and Picha, K. M. (2000) Annu Rev of Biochem 69, 651-697

45. Stano, N. M., Jeong, Y.-J., Donmez, I., Tummalapalli, P., Levin, M. K., and Patel, S. S.

(2005) Nature 435, 370-373

46. Lefebvre, S. D., Wong, M. L., and Morrical, S. W. (1999) J Biol Chem 274, 22830-22838

47. Kreuzer, K. N., and Morrical, S. W. (1994) in Molecular Biology of Bacteriophage T4

(Karem, J., ed), pp. 28-42, American Society for Microbiology, Washington, D.C.

FIGURE LEGENDS

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

13

FIGURE 1. Diagram of bacteriophage T4 proteins and DNA at the replication fork.A, The loop at the fork is formed from 32 protein-covered single-stranded DNA and duplex DNAwith the nascent lagging strand fragment, as suggested by the original trombone model (1). B, Theloop is duplex DNA. The protein-covered single-stranded DNA on the lagging strand is folded into acompact structure, as shown by electron microscopy (2).

FIGURE 2. Replication activity of the biotin-tagged T4 replication proteins. The templatewas nicked pUCnick DNA (2.7 kb). A, Polymerase, B, Helicase, C, N-terminally tagged 59 helicaseloading protein. The products were analyzed by 0.6% alkaline agarose gel electrophoresis and dCTPincorporation determined by trichloroacetic acid precipitation.

FIGURE 3. Replication with biotin-tagged T4 DNA polymerase. Replication products fromnicked pUCnick DNA (2.7 kb) are shown in A, B, D-F. Products of the nicked 452bp mini circle areshown in C. Arrows indicate the biotin DNA pointers showing the location of the tagged polymerase.The pointers were 300 bp in A-E and 179 bp in F. Note that in D-F the leading and lagging strandpolymerases have become separated. In F there are two lagging strand polymerases simultaneouslyelongating fragments, as well as the leading strand polymerase at the fork. The blobs in B and D notassociated with pointers are likely ssDNA segments bound by g32p. G, Number of pointers to biotin-tagged polymerase on different types of replicating molecules, with the pUCnick DNA template.Diagrams depicting molecule type indicate replication products without a rolling circle tail, with asingle-stranded tail (not shown in the micrographs), with a double-stranded tail (not shown) or with 1loop (Fig. 3 A-E) or 2 loops (Fig. 3F) with a double-stranded tail. Scale bar equals 100 nm.

FIGURE 4. Replication with biotin-tagged T4 41 helicase. A and B, Replication products fromnicked pUCnick DNA (2.7 kb). C and D, Replication products from the 452bp mini circle. Arrowsindicate the 300 bp biotin DNA pointers showing the location of the tagged hexameric helicase. Scalebar equals 100 nm. E, Number of pointers to biotin-tagged helicase on pUCnick DNA replicatingmolecules.

FIGURE 5. Replication with biotin-tagged T4 41 helicase and polymerase on stalled andactively replicating molecules. Helicase is present on a smaller fraction of the stalledmolecules. The template was nicked pNNGless on which leading strand synthesis stops after 396 bin the absence of dCTP. A, Biotin-tagged 41 helicase. B, Biotin-tagged polymerase. Arrows indicate300 bp biotin DNA pointers to the tagged helicase or polymerase. Scale bar equals 100 nm. C,Number of pointers to biotin-tagged helicase and biotin-tagged polymerase on stalled and activelyreplicating molecules.

FIGURE 6. Replication with N-terminal biotin-tagged T4 59 helicase loading protein. Thetemplate was nicked pUCnick DNA. Arrows indicate the 300bp biotin DNA pointers to taggedhelicase loader. A-C, Molecules with a compact replication complex at the fork. D-F, Molecules onwhich the complex has opened to show the protein-covered single-stranded DNA on the laggingstrand. Scale bar equals 72 nm for A _ D and 100 nm for E and F. G, Number of pointers to biotin-tagged 59 helicase loading protein on different types of replicating molecules. The diagramsdepicting molecule type indicate replication products without a rolling circle tail, with a single-stranded tail, or with a double-stranded tail, either with or without a loop.

FIGURE 7. Limited binding of biotin-tagged T4 59 helicase loading protein to M13 single-stranded DNA in the absence of replication. Reactions contained M13 viral circular single-stranded DNA (7.2 kb) and all the T4 replication proteins except primase. Arrows show 300bppointers to the helicase loader. Scale bar equals 100 nm.

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from

GriffithNancy G. Nossal, Alexander M. Makhov, Paul D. Chastain II, Charles E. Jones and Jack D.

biopointersArchitecture of the bacteriophage T4 replication complex revealed with nanoscale

published online November 13, 2006J. Biol. Chem. 

  10.1074/jbc.M606772200Access 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

by guest on March 24, 2020

http://ww

w.jbc.org/

Dow

nloaded from