Proc. Natl. Acad. Sci. USA Vol. 86, pp. 9104-9108, December 1989 Biochemistry Escherichia coli replication termination protein impedes the action of helicases Eu Hum LeE*, ARTHUR KornBerG*t, Masumi Hipakat, TAKEHIKO KosayYAsuit, AND TAKASHI Horiucuit *Department of Biochemistry, Stanford University, Stanford, CA 94305-5307; and ‘Department of Molecular Biology, Graduate School of Medical Science, Kyushu University, Maidashi, Fukuoka 812, Japan Contributed by Arthur Kornberg, August 4, 1989 ABSTRACT Identification of the consensus sequence for termination of replication (ter) in Escherichia coli and the isolation of the ter-binding protein (TBP) allowed us to test their effects on replication forks initiated at the unique origin of the E. coli chromosome (oriC) in a purified enzyme system. Rep- lication was severely impeded by fer in a unique orientation when purified TBP was supplied to bind it. The target for blockage within the replication complex can now be ascribed to the inability of dnaB helicase to separate the duplex strands when it encounters ter bound by TBP. Other helicases, such as rep and uvrD proteins, that translocate on DNA and displace strands in the direction opposite to that of dnaB protein are also blocked, but only when the TBP-bound fer is oriented in the other direction. From these results, we infer that the orienta- tion of ter confers a particular polarity on the TBP seated on it, such that a helicase is blocked when it confronts TBP from one side, but can act, presumably by displacing TBP, when facing its other side. Thus, the intrinsic nature of the oriented TBP-¢er complex is responsible for impeding the helicases, rather than any protein-protein interactions. Highly processive DNA polymerases, unlike RNA polymer- ases, continue synthesis on a template strand to the extent that it is available. Template signals comparable to those for termination of transcription that decelerate or block the progress of a replication fork are generally lacking or uni- dentified. Notable exceptions are the termination regions identified in plasmid R6K (1, 2) and the genomes of Esche- richia coli 3-5) and Bacillus subtilis (6, 7). Replication forks moving bidirectionally from the unique origin of each of these genomes are impeded by sequences distantly located around the circular chromosome. Blockage depends on termination (ter) sequences of ~22 base pairs (bp), present as inverted repeats (8), that are nearly identical in R6K and E. coli (5). A dramatic property of the inverted repeats is that one orien- tation blocks fork progress only in the clockwise (CW) direction and the other does so only in the counterclockwise (CCW) direction. An E. coli protein that specifically binds the ter sequences, encoded by the tau (tus) gene, is essential for impeding replication at ter in vivo (9-12). The purified ter-binding protein (TBP; ref. 13 and M.H., T.K., S. Tak- enaka, and T.H., unpublished results) shows the anticipated sequence-specific binding activity. With the fer sequence and purified TBP available, we wished to determine how they operate in combination to impede the progress of a replication fork. In the present study, ter sequences placed in plasmids on either side of oriC (the unique origin of the E. coli chromosome) and bound by TBP exert their orientation-specific blockage of replication by impeding the actions of helicases. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “‘advertisementr’’ in accordance with 18 U.S.C. §1734 solely to indicate this fact. 9104 MATERIALS AND METHODS Reagents. Sources were: unlabeled deoxynucleoside tri- phosphates and ribonucleoside triphosphates (Pharmacia LKB Biotechnology); [y-*2PJATP (6000 Ci/mmol; 1 Ci = 37 GBq) (DuPont/New England Nuclear); [a-?"P]dTTP (800 Ci/mmol) (Amersham); bovine serum albumin (Pentex). Plasmids, Phage DNA, and Synthetic Oligonucleotides. pCM959 (14) is a minichromosome (4012 bp) consisting solely of E. coli DNA encompassing oriC and its flanking genes. M13mp18RF and Mli3mpl19RF DNA were from New En- gland Biolabs. Synthetic oligomers were made by K. Relloma (Stanford University) using a DNA synthesizer (Coder 300, DuPont). Enzymes and Proteins. T4 polynucleotide kinase, T4 DNA ligase, and Hae Ill, HindIII, EcoRV, Pst 1, SacI, Sac Il, and Xho I were purchased from New England Biolabs. UvrD protein was a gift from Paul Modrich (Duke University, Durham, NC). Rep protein and highly purified replication proteins were prepared as described (15, 16). TBP was overproduced 2000-fold in cells carrying a plasmid with mutations at the promoter site of the tau gene. The protein was purified from cell extracts to apparent homogeneity by DEAE-Sephacel and Heparin-Sepharose CL-6B chromatog- raphy (M.H. et al., unpublished results). Reconstituted oriC DNA Replication. The reaction mixture (40 yl) contained 30 mM Hepes‘KOH (pH 7.6), 8 mM magnesium acetate, 2 mM ATP, template DNA (600 pmol of nucleotide), 0.3 mM CTP, 0.3 mM UTP, 0.3 mM GTP, 0.1 mM dGTP, 0.1 mM dCTP, 0.1 mM dATP, 0.01 mM [a- 32P}TTP (100-200 cpm/pmol), 150 mM potassium glutamate, 100 ng of dnaA, 700 ng of SSB, 60 ng of dnaB, 20 ng of dnaC, 50 ng of gyrase A, 10 ng of gyrase B, 8 ng of HU, 10 ng of primase, 70 ng of DNA polymerase III holoenzyme, and 75 ng of B subunit of DNA polymerase holoenzyme. After incubation at 30°C, reactions were stopped by adding EDTA to 20 mM. The extent of DNA synthesis was measured by liquid scintillation counting after precipitation with 10% (wt/vol) trichloroacetic acid and filtration onto Whatman GF/C glass-fiber filters. The progress of bidirectional repli- cation was assayed in stages. The prepriming complex was formed in 20 yl containing 40 mM Hepes-KOH (pH 7.6), 15% (vol/vol) glycerol, bovine serum albumin (200 ug/ml), 0.007% Brij 58, 0.25 mM EDTA, 200 mM potassium gluta- mate, 3.3 mM ATP, template DNA (600 pmol of nucleotide), 8 ng of HU, 60 ng of dnaB, 20 ng of dnaC, and 100 ng of dnaA. After 15 min at 37°C, the temperature was shifted to 18°C for 2 min, followed by addition of the missing replication com- ponents (see above) in a final volume of 30 ul that also contained magnesium acetate at 10 mM. Incubation was continued for 2, 4, or 6 min; the reactions were stopped by adding EDTA to 20 mM. A small portion of each sample was used to determine DNA incorporation as described above. Abbreviations: TBP, ter-binding protein, CW, clockwise; CCW, counterclockwise. +To whom reprint requests should be addressed. Biochemistry: Lee er al. Samples were extracted twice with phenol, once with phenol/ chloroform (1:1), twice with chloroform, and then precip- itated with ethanol. The DNA pellets were redissolved in buffer containing 10 mM Tris-HCl (pH 8.0) and 0.5 mM EDTA and digested with Hae III and EcoRV as directed by the supplier. Agarose and Polyacrylamide Gel Electrophoresis. Essen- tially as described in ref. 17. Samples (10-20 pl) of the replication reaction were made up to 0.5% SDS and loaded on a 0.8% alkaline agarose gel (in 50 mM NaQH/1 mM EDTA). The gels, electrophoresed for 16 hr at 40 V, were neutralized by soaking in 1.5 M Tris-HCl (pH 7.5), dried, and subjected to autoradiography. Restriction endonuclease (Hae III and EcoRV) digests of replicated DNA were brought to 0.5% SDS and loaded on a 7% polyacrylamide gel (acrylamide/ N,N'-methylenebisacrylamide ratio 30:1) containing 10% glycerol, 37 mM Tris borate (pH 8.3), and 1 mM EDTA. The gels, electrophoresed for 4 hr at 10 V/cm, were dried, autoradiographed, and scanned by densitometry. Preparation of the DNA Helicase Substrates. The 244-bp HindHil-Pst 1 fragment containing the ter sequence was excised from oriC ter CW and inserted into the HindIII-Pst I cleaved site of M13mp18 DNA (duplex replication form) from which the viral, single-stranded circular M13mp18 ter CCW was generated (18). A 10-fold excess (molar ratio) of a 30-mer oligonucleotide containing the ter sequence [5’-end labeled with T4 polynucleotide kinase (17)] was annealed to Ml3mp18 ter CCW in 10 mM Tris-HCI (pH 8.0), 100 mM NaCl, and 1 mM EDTA at 100°C for 2 min and cooled slowly to 50°C. Incubation was continued for 1 hr at 50°C, and then samples were cooled to room temperature. The template, with the 30-mer oligonucleotide annealed to it, was purified by a Bio-Gel A-5m spun column (17), extracted with phenol and chloroform and then precipitated with ethanol. Similarly, a HindIW-Pst I-cleaved 244-bp fragment of oriC ter CW was inserted into the HindIII-Pst I site of M13mp19 to generate a 28-bp duplex with M13mp19 ter CW. The 5’-end labeling of the 28-mer oligonucleotide, annealing, and purification pro- cedures were as above. Assays of Helicases: rep, uvrD, and dnaB. Assays of rep and uvrD were as described (19, 20). In brief, the reaction mixture (20 yl) contained 40 mM Tris-HCl (pH 7.6), 5 mM dithio- threitol, 2.5 mM MgCh, bovine serum albumin (50 «xg/ml), 2.5 mM ATP, and 0.12 pmol (as circles) of the DNA sub- strate. The amounts of rep, uvrD, and TBP were as indicated. The reaction was incubated at 37°C for 10 min and terminated by adding an equai volume of a solution containing 80 mM EDTA, 1% SDS, 20% glycerol, 0.02% bromophenol blue, and 0.02% xylene cyanol. Samples were loaded on a nondena- CACTTSAGT TACKS TCGAGTGAAATCAATET > OTRTGTTGTAACTAAAGTAT 3 BG LAGATAGATT TGATAGC Proc. Natl. Acad. Sci. USA 86 (1989) 9105 turing 10% polyacrylamide gel, dried, and subjected to au- toradiography. Slices of the dried gel that contained the | undisplaced (substrate) and displaced (oligonucleotide) bands were cut out; from the radioactivity measured in a liquid scintillation counter, the percentage of displaced oli- gonucleotides was calculated. The dnaB helicase assay was as described (21). The mixture (20 yl) contained 40 mM Hepes‘KOH (pH 7.6), 11 mM magnesium acetate, 3.4 mM ATP, bovine serum albumin (50 ug/ml), and 0.12 pmol of DNA circles. The amounts of TBP and dnaB were as indi- cated. The procedure was the same as for rep and uvrD. Preparation of Fraction II. E. coli W3110 cell paste was lysed as described (22). The supernatant (fraction I) was precipitated with ammonium sulfate (0.26 g/ml). The centri- fuged pellet was resuspended in buffer containing 20% glyc- erol, 50 mM Hepes‘KOH (pH 7.6), 100 mM NaCl, 1 mM dithiothreitol, and 1 mM EDTA. RESULTS Construction of oriC ter Plasmids. Segments containing the 22-bp ter sequence were inserted in the 4-kilobase plasmid (pCM959) that contains oriC, the minimal 245-bp replication origin of the E. coli chromosome (Fig. 1). In one plasmid (oriC ter CW), ter was placed to the right of orviC in an orientation expected to block the replication fork progressing CW from oriC. In another plasmid (oriC ter CCW), ter was placed to the left of oriC in an orientation that should block CCW replication. The orientations were based on the polarity of the sequences observed in vivo (2, 5, 11). The copy numbers of the ter CW and ter CCW plasmids appeared to be 60% that of the oriC plasmid. On the other hand, plasmids in which ter sequences were placed on both sides of oriC (oriented to prevent both CW and CCW replication) were present at only 5—10% the copy number of the oriC plasmids. Thus the capacity for autonomous replication of an oriC plasmid is severely reduced when the progress of replication forks is impeded in both directions. Replication of oriC ter Plasmids Is Inhibited. With the purified enzyme system, replication of the oriC ter plasmids, compared to that of oriC was only slightly less active (Fig. 2). However, upon the addition of a crude E. coli enzyme fraction (fraction II), the inhibition of replication reached 40% (data not shown). The factor in the crude fraction that elicits the ter activity could be replaced by purified TBP. Saturation for the inhibition of replication was achieved at about an equimolar ratio of TBP to fer. In the replication of oriC plasmids, the initial products are genome-length strands; in an ensuing mode of replication, the Fic. 1. Construction of oriC ter CCW and oriC ter CW plasmids. Two complementary synthetic oligomers of 30 and 28 nucleotides were annealed to generate a 26-bp duplex containing the 23-bp ser consensus sequence and restriction nuclease Sac I (SI) and Sac II (SH) cohesive ends. oriC plasmid DNA (pCM959, 4012 bp) was cleaved by Sac I and Sac IH to generate a site for insertion of the synthetic duplex. In the resultant oriC ter CCW plasmid circle (3812 bp), the ter sequence is oriented to block CCW DNA replication. In a similar way, fer in a 26-bp sequence with cohesive ends for Xho I (X) and Cla I (Cl) was inserted into an oriC plasmid cleaved by the same enzymes. In the resultant oriC ter CW plasmid (4028 bp), the rer sequence is oriented to block the CW DNA replication fork. The T-like symbol representing ter is oriented such that the head of the T blocks the progress of the replication fork, based on the polarity observed in vivo (2, 5, 11). The cleavage sites (in bp) in oriC are: Cla 1, 426; Sac 1, 3633; Sac U1, 3849; and Xho I, 417. 9106 Biochemistry: Lee et al. 800 PLASMID TBP a offic - ~ 0 ofc + £ 600 @ oriC-terCW = & 4 orfC- terCCW - a a wi 400 E z >» oa < Z 200 o oriC-terCW + 4 orlC-terCCW + 0 l 1 j 0 5 10 15 20 25 TIME (min) Fic. 2. TBP inhibits replication of oriC ter plasmids. TBP (72 ng) was added where indicated (+). -, TBP not added. products are strands of multimer length. Analysis of the distribution of monomer- and multimer-length products by alkaline agarose gel electrophoresis showed that the inhibi- tion of replication observed with the ter plasmids was largely at the presumed rolling-circle stage of product amplification (Fig. 3). The result is explicable in view of the bidirectional mode of oriC plasmid replication (23, 24). When the CW direction is blocked by oriC ter CW, a unit-length strand can still be produced in the CCW direction; similarly, blockage of the CCW direction permits a unit-length strand to be gener- ated in the CW direction. Multimer-length strand synthesis, requiring repeated traversals of the entire genome, is the stage most affected by the TBP-ter complex as manifested in the total DNA synthesis. 150 A) MONOMER a PLASMID TBP eoriC-terCcW - 100 sonic - oorlc + aoriC>terCCW - oorC-terCW + AorlC:terCCW + _ 50 3 —E 3s a a 0 z z B) MULTIMER PLASMID JBP_ % 600 |- a oriC> terCCW - s s a - 0 ofl + a . © offC+terCW - 400 }- eo 200 — a o orlC-terCW + 4 oriC-terCCW + | } I 1 Oc 0 10 20 30 40 50 60 TIME (min) Fic. 3. Stage in oriC ter plasmid replication inhibited by the TBP-ter complex. Products of replication (reconstituted system), containing about 10 ng of TBP where indicated (+), were subjected to alkaline agarose gel electrophoresis and autoradiography. Radio- activity in slices of the dried gel that contained the monomer-length (A) and multimer-length (B) DNA bands was measured in a liquid scintillation counter. Each value represents the average of two experiments. ~, TBP not added. Proc. Natl. Acad. Sci. USA 86 (1989) Replication Block Is Located at or Near the ter Site. The progress of replication of oriC and oriC ter CW plasmids, with or without TBP, was monitored by the sequential appearance of radioactive nucleotides in fragments of the plasmid generated by Hae III and EcoRV cleavage (Fig. 4). Based on the amount of label in each fragment (corrected for its size), as a percentage of the label in all fragments, replication of the oriC plasmid proceeded bidirectionally, starting at or near oriC as reported (23, 24); the presence of TBP had no effect on the pattern. However, the pattern of oriC ter CW plasmid replication differed strikingly when TBP was present. Replication was unidirectional, moving leftward (CCW) because the TBP-ter CW complex blocked its CW movement. The same experiment, repeated with oriC ter CCW, produced a pattern that was virtually the mirror image of that observed with oriC ter CW; replication proceeded only CW from oriC (data not shown). TBP Impedes the Actions of Helicases. Progress of replica- tion of duplex DNA requires a separation of the strands to provide the template and depends on the action of a helicase as the leading element. In the case of an oriC plasmid, once it is primed, replication by DNA polymerase III holoenzyme depends on the actions of dnaB protein and gyrase (24, 25); the rate of fork movement is determined by the rate of unwinding. To assess how dnaB protein and the helicases are affected by ter, substrates were constructed in which 30-bp Lo L149 18 L7 LE £5 14 L3 L2 14 Ra RI R2 R3 ™ SK LL Ye | i | Petro re Pl ont 20 ont ort + TBP ori @ ter CW LABEL IN FRAGMENT (% of total) oe oriC @ terCW + TBP 8 1 2 3 4 LENGTH (kb) Fic. 4. TBP-ser complex blocks progress of the replication fork in one direction. The oriC reconstituted system was staged and polyacrylamide gel electrophoresis and autoradiography were per- formed. The autoradiogram was scanned by densitometry and each peak on the graph was cut out and weighed. The density values were corrected for fragment size. The values shown ate for a 6-min replication. DNA fragments were generated by EcoRV and Hae Ill digestion; LO to L10 represent fragments in CCW order from oriC; RO to R3 are in CW order. Sizes (in bp): RO, 785; R1, 250; R2, 926; R3, 216; LO, 785; L1, 108; L2, 9; L3, 448; L4, 484; LS, 76; L6, 85; L7, 374; L8, 22; L9, 45; L10, 204. The positions of the minimal oriC and ter (T-shaped area in LO/RO) are shown. kb, Kilobases. Biochemistry: Lee ef al. “rap. werd ena’ (b) Fic. 5. Substrates for the helicases. Two substrates were con- structed. (a) In M13mp18 ter CCW; the ter sequence (T-shaped area) is oriented to block CCW replication. (6) In Mi3mp19 ter CW, the ter sequence (T-shaped area) is oriented to block CW replication. The orientations of ter sequences were confirmed in each case by use of the M13 dideoxynucleotide sequencing method (26). oligonucleotides were annealed to single-stranded circles to generate the duplex ter sites in either orientation (Fig. 5). Upon exposure to a helicase, displacement of the oligonu- cleotide was measured electrophoretically. Three helicases were examined: dnaB protein, which translocates on a DNA strand in the 5’—3' direction (CW) (21), and rep (19) and uvrD (20) proteins, which translocate in the 3‘—>5' direction (CCW) (Fig. 5). Strand displacement by each of the helicases was observed in the absence of TBP, showing that the ter sequence as such offers no impediment (Figs. 6 and 7). The dnaB protein that prefers to displace a strand with an unannealed 3’ end (e.g., a ‘‘tail’”) (21) was still sufficiently active on this fully annealed (‘‘tailless’’) oligo- nucleotide. In contrast, in the presence of TBP, dnaB protein failed to displace the strand specifically when fer was ori- ented to block CW replication, whereas rep and uvrD pro- - teins were impeded in their actions by ter only in the orientation that blocks CCW movement (Figs. 5-7). Thus, TBP binds ter in both orientations, but blocks the action of a helicase only in one direction, the same polarity as that observed for blockage of a replication fork in vivo. The appearance of the displaced oligonucleotide as two bands in the native gel electrophoresis (Fig. 6) in all the samples (including the heat-treated substrate) is likely due to the formation of some secondary structure or to dimerization, possibly through annealing at a potential hairpin loop in the middle of ter. Electrophoresis in a denaturing urea gel revealed the oligonucleotide in all samples as a single band in the position expected for its size. DISCUSSION In recent years, knowledge regarding the termination of replication has been enlarged by two major advances: (i) recognition that a defined sequence (ter), present as inverted repeats (opposite orientations), is responsible for termination by impeding the bidirectional replication of plasmid (1, 2), E. coli (5, 11), and B. subtilis (6) genomes; and (ii) isolation of TBP, the product of the tau or tus gene, which in a complex with ter is essential for the termination event (7, 10, 12, 13). With a system of purified proteins that carries out the bidirectional replication of plasmids initiated at oriC (the unique origin of the E. coli chromosome) (16, 24), we have been able to show that both ter and TBP are necessary and sufficient for reconstitution of termination in vitro (Figs. 2 and 3). As expected, one orientation of ter impedes replica- tion fork movement only in a CW direction (Fig. 4), whereas the other orientation inhibits it only in the CCW direction. As further confirmation, replication of plasmids with the ter sequence on both sides of oriC is nearly completely blocked (E.H.L. and A.K., unpublished results). Inasmuch as fork movement in the reconstituted oriC plasmid replication system depends on the action of a heli- case to separate the strands of the duplex, the behavior of Proc. Natl. Acad. Sci. USA 86 (1989) 9107 PLASMID }-M13mp18+terCCW df 1 3mp 19-ter CW HEAT HELICASE 0 oa o> o[ TBP -— me me ~— = + $e eH OH 123465 ~_ + + 6 7 8 9 10111213 14151617 rep uyvrD dnaB ‘ fe "? { 123 45 6 7 8 9 10111213 14151617 Fic. 6. TBP-ter complex inhibits helicases in an orientation- specific manner. The products of the reactions were separated by 10% polyacrylamide gel electrophoresis and detected by autorad- iography. In lane 17 the substrate was heated at 100°C for 2 min. Helicases tested were rep (Top), uvrD (Middle), and dnaB (Bottom). As indicated, the helicase protein was omitted in lanes 1, 5, 9, and 13 and present in increasing amounts in lanes 2-4, 6-8, 10-12, and 14-16. The increasing amounts of protein in the three-lane groups above were 4, 17, and 53 ng for rep protein; 14, 41, and 125 ng for uvrD helicase; and 26, 150, and 272 ng for dnaB helicase. The corresponding molar ratios of helicase to substrate were 0.5, 2, and 7 for rep; 2, 5, and 14 for uvrD; and 0.7, 4, and 7 for dnaB. helicases encountering a TBP-rer complex was examined (Fig. 5). Each of three helicases was strikingly impeded in the separation of strands at a fer site bound by TBP. The dnaB protein, which translocates on DNA ina 5'—>3’ direction, was blocked by the TBP-ter complex only when oriented to impede CW fork movement, whereas the actions of rep and uvrD proteins, which translocate on DNA in the 3’>5' direction, were inhibited by the TBP-ter complex only in the orientation that blocks CCW movement (Figs. 6 and 7). Even though one or another helicase may be a major element in 9108 Biochemistry: Lee et al. 100 Proc. Natl. Acad. Sci. USA 86 (1989) rep (TBP) CWw(+) DISPLACED OLIGOMER (percent) uvrD (TBP) dnaB (TBP) Cwl-) CW(+) ccwl-) CCW(+) rep (ng) 80 uvrD (ng) 80 160 dnaB (ng) Fic. 7. Slices of the gel in Fig. 6 containing the displaced and undisplaced oligomers were analyzed for radioactivity. determining the rate of fork movement, the influence of the TBP-ter complex on other constituents of the massive ‘‘rep- lisome”’ assembly of polymerases, primosome, and binding proteins will have to be considered as well. Plausibly, the obstruction by the TBP-rer complex to the actions of these three helicases applies to other helicases as well. Specific protein-protein interactions are not indicated as a basis for the blockage. In strand displacement by E. coli DNA polymerase I, a polymerase that can act without ATP hydrolysis or assistance of a helicase, both orientations of the complex were inhibitory to the large fragment (Klenow fragment), but had less effect on the intact enzyme (E.H.L. and A.K., unpublished results). Thus, the polymerase with- out the energy input of an ATP-driven helicase can overcome the obstacle posed by TBP, but the partial polypeptide of the polymerase is less well equipped to do so. The basis for the polarity of the TBP-ter complex in blocking strand separation needs to be examined. Clearly, TBP bound asymmetrically to ter, when faced in one direc- tion, is an impediment to strand separation but, when ap- proached from the opposite face, presents no obstacle and is presumably displaced. Structural and footprinting studies of the complex in the course of helicase actions may reveal distinctions between the two orientations of TBP bound at a ter site. Further work is also needed to determine the degree to which the TBP-ter complex promotes pausing or even the dissociation of otherwise processive helicase and replication proteins. The value of a termination mechanism to a replicon is uncertain. Most genomes (e.g., ColE1 plasmid, phage A, and simian virus 40) lack such a mechanism and those that possess it suffer no serious disadvantage when fer sites are removed or the TBP is inactivated by mutation. Perhaps the topological states generated by replication forks as they approach each other or approach the origin of replication may offer difficulties, such as a susceptibility to a rolling-circle mode of replication, that are resolved by slowing the pace of strand separation at the advancing fork. The temporally regulated replication of large chromosomes may be another instance in which a termination mechanism is employed. This work was supported by grants from the National Institutes of Health and the National Science Foundation to Stanford University and for scientific research at Kyushu University from the Ministry of Education, Science, and Culture, Japan. 1. 2. 3. “ ye NS 11. 12. 13. 14. 15. 16, 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Bastia, D., Germine, J., Crosa, J. H. & Ram, J. (1981) Proc. Natl. Acad. Sci. USA 78, 2095-2099. Horiuchi, T. & Hidaka, M. (1988) Cell 54, 515-523. de Massy, B., Louarn, S. B., Louarn, J. M. & Bouché, J. P. (1987) Proc. Natl. Acad. Sci. USA 84, 1759-1763. Pelletier, A. J., Hill, T. M. & Kuempel, P. L. (1988) J. Bac- teriol. 170, 4293-4298. Hidaka, M., Akiyama, M. & Horiuchi, T. (1988) Cell 55, 467-475. Lewis, P. J. & Wake, R. G. (1989) J. Bacteriol. 171, 1402-1408. Lewis, P. J., Smith, M. T. & Wake, R. G. (1989) J. Bacteriol. 171, 3564-3567. Carrigan, C. M., Haarsma, J. A., Smith, M. T. & Wake, R. G. (1987) Nucleic Acids Res. 15, 8501-8509. Hill, T. M., Kopp, B. J. & Kuempel, P. L. (1988) J. Bacteriol. 170, 662-668. Hill, T. M., Tecklenburg, M. L., Pelletier, A. J. & Kuempel, P. L. (1989) Proc. Natl. Acad. Sci. USA 86, 1593-1597. Pelletier, A. J., Hill, T. M. & Kuempel, P. L. (1989) J. Bac- teriol. 171, 1739-1741. Kobayashi, T., Hidaka, M. & Horiuchi, T. (1989) EMBO J. 8, 2435-2441. Sista, P. R., Mukherjee, S., Patel, P., Khatri, G. S. & Bastia, D. (1989) Proc. Natl. Acad. Sci. USA 86, 3026-3030. Meijer, M., Beck, E., Hansen, F. G., Bergman, H. F., Messer, W., von Meyenberg, K. & Schaller, H. (1979) Proc. Natl. Acad. Sci. USA 76, 580-584. Scott, J. F. & Kornberg, A. (1978) J. Biol. Chem. 253, 3292- 3297. Kaguni, J. M. & Kornberg, A. (1984) Cell 38, 889-900. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Labora- tory, Cold Spring Harbor, NY). Messing, J. (1983) Methods Enzymol. 101, 20-78. Smith, K. R., Yancey, J. E. & Matson, S. W. (1989) J. Biol. Chem. 264, 6119-6126. Wood, E. R. & Matson, S. W. (1987) J. Biol. Chem. 262, 15269-15276. LeBowitz, J. M. & McMacken, R. (1986) J. Biol. Chem. 261, 4738-4748. Shlomai, J. & Kornberg, A. (1980) J. Biol. Chem. 255, 6789- 6793. Kaguni, J. M., Fuller, R. S. & Kornberg, A. (1982) Nature (London) 296, 623-627. Baker, T. A., Funnell, B. E. & Kornberg, A. (1987) J. Biol. Chem. 262, 6877-6885. Baker, T. A., Sekimizu, K., Funnel], B. E. & Kornberg, A. (1986) Cell 45, 53-64. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.