DNA replication: enzymology and mechanisms Zvi Kelman and Mike O'Donnell

Cornell University Medical College, New York, USA

Research into the enzymology of DNA replication has seen a multitude of highly significant advances during the past year, in both prokaryotic and eukaryotic systems. The scope of this article is limited to chromosomal replicases and origins of initiation. The multiprotein chromosomal replicases of prokaryotes and eukaryotes appear to be strikingly similar in structure and function, although future work may reveal their differences. Recent developments, elaborating the activation of origins in several systems, have begun to uncover mechanisms of regulation. The enzymology of eukaryotic origins has, until now, been limited to viral systems, but over the past few years, enzymology has caught a grip on the cellular origins of yeast.

Current Opinion in Genetics and Development 1994, 4:185-195

Introduction In this review, we will focus on recent developments in the study of origin activation and the mechanism of The structure of duplex DNA is so simple and elegant action of chromosomal replicases. These areas encom- that one may have thought it would be simple to du- pass only a subset of the recent advances in the enzy- plicate. However, since the isolation of DNA polymer- mology of DNA replication. ase I first delivered replication mechanisms into the hands of enzymologists, the process has been shown to be far from simple. Over 20 are utilized in Es- cherichia coil and probably more in higher organisms [1]. In fact, the individual functions of many of these Replication origins proteins are still unknown and even more proteins re- main to be identified. In overview, the process begins The first (and still the only) cellular origin to have at a specific sequence called an origin upon which been activated in vitro was oriC, the origin of the proteins bind and locally unwind the duplex, allow- E. coil [1]. Since that breakthrough, sev- ing invasion of a . The helicase couples ATP eral phage and viral origins have been activated in hydrolysis to melt the duplex, and then the single- vitro. Origin activation requires multiple copies of stranded DNA (ssDNA) is coated with ssDNA-bind- an origin-binding and usually additional ori- ing protein (SSB). This SSB-ssDNA nucleofilament has gin accessory proteins, some of which have secondary no secondary structure and serves as the most efficient functions as transcriptional activators. As proposed by template for chromosomal replicases, multiprotein ma- Hatch Echols [2], the requirement for multiple proteins chineries characterized by their rapid and highly pro- in site-specific processes, such as activation of origins cessive DNA synthesis. One strand of the chromosome and promoters, is imposed by the need for a high fi- is synthesized continuously, but because of the antipar- delity of action on specific sequences embedded in the allel structure of the duplex, the other strand is copied bulk of chromosome DNA. These multiprotein-DNA discontinuously as a series of fragments (Okazaki frag- complexes are termed specialized nucleoprotein struc- ments). This discontinuous mode of replication neces- tures (snups) [2]. Origin initiation events in E. coli, sitates frequent reinitiation of DNA chains, which are phage ~ and simian virus 40 (SV40) are similar and primed by short RNA primers synthesized by a pri- appear to apply also to the more recently characterized mase. Most of the above proteins are thought to work systems of phage P1, bovine papilloma virus (BPV), together in one large assembly for coordi- herpes simplex virus 1 (HSV-1), and yeast (see Fig. nated synthesis of both strands of the chromosome. 1). In most systems, the origin-binding protein is pre-

Abbreviations ACS~ARS core consensus sequence; ARS.--autonomous replication sequence; BPV--bovine papilloma virus; HSV-l--herpes simplex virus 1 ; MCM--minichromosome maintenance; ORC---origin recognition complex; PCNA--proliferating cell nuclear antigen; pol/PoI~DNA ; PP2A--protein phosphatase 2A; RF-C--; RP-A--; snup---specialized nucleoprotein structure; SSB~single-stranded DNA-binding protein; ssDNA--single-stranded DNA; SV40--simian virus 40.

© Current Biology Lid ISSN 0959-437X 185 186 and expression mechanisms

pared for origin activation and, for this review, we de- late the process is not yet clear. In stage III, the ori- fine this as stage I. Multiple copies of the origin-binding gin snup unwinds DNA within three AT-rich 13-mer protein then assemble onto the origin with the acces- repeats to form the open complex and this reaction sory proteins to form the snup (stage II). This untwists requires a supercoiled template [1]. Formation of the a section of AT-rich DNA to form an 'open complex' open complex is highly regulated [1]. Negative regula- (stage III), providing a point of entry for the helicase tion is achieved by the inhibitor of cellular initiation, (stage IV). After helicase entry, the chromosome can IciA, which competes with DnaA for interaction with be extensively unwound and replication forks can be the 13-mers and prevents the formation of the open assembled. Origin activation systems differ in the de- complex [7"°,8"°]. ATP also regulates the open com- tails of these steps and also in their regulation. In the plex formation. DnaA can bind either ATP or ADP following section, we will focus on these differences, tightly, but only the ATP form is active in unwinding with particular attention to recent developments. the 13-mers [11. The DnaA protein becomes inactive upon slowly hydrolyzing the ATP to ADP, and the ADP remains bound, thus preventing subsequent reinitia- Prokaryotic origins tions. The DnaA protein can be readily reactivated by Recent studies show that the activation of oriC is reg- acidic phospholipids, which catalyze the exchange of ulated at each stage (Fig. 1). In stage I, the DnaA ATP for the bound ADP [1,9"'], but only when DnaA origin-binding protein is inactive when bound to acidic is bound to or/C, this may underlie the early obser- phospholipids, but this repression can be relieved, and vations of attachment to the membrane [10]. DnaA prepared for origin binding, by the DnaK heat RNA polymerase provides yet another level of reg- shock protein or by phospholipase treatment [3]. In ulation in determining the ability of R loops (RNA stage II, the origin snup is nucleated at four DnaA- paired with one strand of the DNA duplex) to pro- binding sites in oriC, upon which 20 or more DnaA mote the open complex, and although the transcript monomers assemble, along with several origin acces- need not enter on'C, it must be close [11]. In stage IV, sory proteins [1]. These accessory proteins include HU the DnaB helicase enters the open complex, but its ac- and IHF, which bend DNA, and they may act by help- cess absolutely requires the DnaC protein, a molecu- ing to wrap DNA around the DnaA subunits [4°',5]. lar matchmaker that delivers the hexameric DnaB into Other accessory proteins include Fis and the recently the origin in an ATP-dependent reaction, after which identified Rob protein [6"°]. They are not essential for DnaC departs from the DNA [1]. Origin unwinding is the activation of the origin, and how they may modu- bidirectional; therefore, two hexamers of DnaB must

Stage I Stage I1 Stage III Stage IV DNA ~ r-f Open

Prepare origin-binding Open complex Origin- protein ~ formation binding | I~ Replication fork System protein II + - II II + - II Helicase proteins E. coil DnaA DnaK Phospholipids HU. IHF, ATP ADP DnaB Dna(- Excess'~ Phospholipase Fis, Rob Acidic phospholipids IciA DnaC --| R loops / Dna(;-Prima~e Phage ~. ~. O HU RNA pol DnaB ~. p ~. p ~ SSB Dnal. DnaK ~ DNA po] III hoh,enzyme Gq)E Phage PI RepA DnaJ, DnaK, DnaA DnaB GrpE

Yeast ORC ATP, ABFI -~ i)okx-Prin'=ase SV40 T-antigen Cdc2 psi ATP banligen RP-A(SSBI p53 ~ RP-A(SSB] PP2A Spl PP2A pol 8, t" (+RF-C, PCNA) BPV E1 E2 E!

HSV-I UL9 UL5,8,52 ICP8 Helle ase--(UL5,8.521 ICP8 (SSB) HSV-1 Pol q+LJt42) @1994 Currenl Opinion In Genetics and Developmenl

Fig. 1. Regulation at different stages of origin activation in a number of systems. Origin activation has been subdivided into four stages: preparation of the origin-binding protein, assembly of the origin snup, local unwinding of origin DNA (the open complex), and entry of the helicase. Molecules that act on each stage are divided into positive (+) and negative (-) effectors and are shown for E. colt, phage )~, phage P1, yeast, SV40, BPV and HSV-1. A blank entry indicates that effectors of that step have not (yet) been identified. At the far right are listed the replication proteins needed to advance the replication fork. DNA replication: enzymolo~/and mechanisms Kelman and O'Donnell 187 be delivered to the open complex. Negative regula- the origin snup, is facilitated by ATP binding, which tion is achieved by the presence of too much DnaC effects the assembly of two hexamers of T-antigen protein, which binds to and inhibits the DnaB helicase onto the origin (reviewed in [25]). The two hexam- activity, possibly by preventing its translocation along ers are thought to encircle the origin DNA in SV40, DNA; therefore, a 'fine balance' of DnaC and DnaB is because if the hexamers are pre-assembled in solution essential to the productivity of this stage [12]. they do not bind the origin [26"']. Moreover, in elec- Phage ~. uses mainly host replication proteins for origin tron micrographs the double hexamer of T-antigen on activation, with the exception of the phage-encoded O the origin does not appear to wrap DNA around it as and P proteins, analogs of DnaA and DnaC [1]. The ;k with other origin snups, but rather the DNA appears origin snup is composed of four dimers of ~. O pro- to follow a straight path through the protein [27]. The tein, which do not require pre-activation to assem- transcriptional activator Spl acts as an origin accessory ble onto the origin. Formation of the open complex protein that perturbs the local histone distribution to at three AT-rich l 1-met repeats requires the DNA to make the origin accessible for initiation proteins [28]. be supercoiled, but does not require ATP [13,14]. In However, the identity of the transcriptional activator stage IV, the ~. P protein binds DnaB and delivers it is not important, as Spl can be exchanged for other to the origin, but the P protein remains tightly associ- activators [28], but the DNA-binding domain must be ated with DnaB and prevents its helicase activity [15]. accompanied by an activation domain for origin func- In this regard, the P protein acts as a negative regulator, tion [29"']. like excess DnaC, to prevent translocation of DnaB. Al- Formation of the open complex in stage III is coinci- though ATP is not required for P protein to bring DnaB dent with T-antigen binding to the origin, which results to the origin, it is required to break the interaction in the melting of 8 bp in the early palindrome and un- of P protein with DnaB, thus freeing the helicase for twisting of an AT tract [25]. These events require bind- replication; this step is mediated by heat shock pro- ing, but not hydrolysis, of ATP and do not require a teins DnaJ, DnaK and GrpE [16,17]. Transcription from supercoiled template [25]. In SV40, unlike the other sys- the rightward promoter is needed for ~. origin activa- tems discussed above, the stage IV helicase invasion is tion in vivo, and in vitro studies show that transcription unique in that T-antigen itself is the helicase and thus disrupts HU-mediated negative repression (presumably encompasses the functions of the E. coli DnaA, B and of stage II or III) [18]. C proteins. Stage IV is still regulated, however; for ex- Recent studies on the activation of the lysogenic origin ample, it has been shown recently that PP2A-treated of phage P 1 show that the phage-encoded RepA origin- T-antigen provides a cooperative interaction between binding protein is a native dimer, but the monomer the two hexamers, which facilitates stage IV [30"]. In strongly associates with the origin. In stage I, RepA addition, the human SSB, replication protein A (RP-A), is prepared for origin binding by the DnaJ, DnaK and consists of three non-identical subunits (p70, p32 and GrpE heat shock proteins, which couple ATP hydroly- p14) and has been shown to be a positive effector of sis to monomerize RepA [19"']. The origin snup com- stage IV, as RP-A is needed for T-antigen to unwind prises RepA bound at five sites and the host DnaA pro- the bulk of DNA [25]. A second point at which p53 tein at five other sites. At least two of the DnaA-binding may regulate replication has been identified recently sites are essential for origin function, and in vitro stud- in an interaction of p53 with the p70 subunit of RP-A ies show that the ADP form of DnaA protein is capable (the ssDNA-binding subunit) that inactivates its ability of activating the origin in combination with RepA (re- to bind ssDNA [31"']. Also, the p32 subunit of RP-A is viewed in [20]). Although stages III and IV have not phosphorylated in a cell cycle dependent manner in been studied in detail, it seems likely that the open the G1 to S phase transition by members of the cy- complex forms at five slightly AT-rich 7-met repeats clin cdc2 kinase family, and addition of this kinase to and then DnaC mediates delivery of DnaB to the open the SV40 replication system stimulates DNA synthesis complex. [32"']. Phosphorylation of p32 is also performed by the DNA-activated protein kinase (GS Brush, CW Ander- son, TJ Kelly, abstract 191, Eukaryotic DNA Replication Eukaryotic viral origins Meeting, Cold Spring Harbor, September 1993). In the SV40 system (reviewed in [21]), the stage I prepa- The BPV snup is composed of two viraUy encoded ration of the vitally encoded T-antigen for origin bind- proteins, E1 and E2. E1 is the origin-binding protein ing is regulated in both positive and negative fash- and a helicase (like T-antigen) [33",34"'] and E2 is an ion. T-antigen must be phosphorylated at Thr124 for origin accessory protein (and transcriptional activator) productive origin binding, and this modification can that binds specific sequences and helps in the delivery be performed by the cdc2 kinase [22]. Phosphoryla- of E1 to the origin via direct protein contacts [35",36]. In tion at other sites can inhibit T-antigen, and protein this regard, E2 fulfills a stage IV function analogous to phosphatase 2A (PP2A) has been shown to activate that of E. coli DnaC in delivery of the helicase into the T-antigen by removing phosphates from specific ser- origin. Regulatory aspects, exact functions of ATP and ine residues implicated in DNA-binding activity [23]. identification of an open complex remain for future The p53 tumor suppressor protein is a negative reg- studies, but the in vitro replication system and avail- ulator of stage I; it binds to T-antigen preventing its ability of pure E1 and E2 should yield this information association with the origin [24]. Stage II, formation of in the near future. 188 Chromosomesand expression mechanisms

In the HSV-1 system, the identification of all seven of S phase [44",45"']. It is hypothesized that in the nucleus the essential virally encoded replication proteins and the MCM proteins activate the ARS and are then trans- the cloning and production of each of them in quan- ferred to the cytoplasm to prevent reinitiation [44",45"']; tity has enabled a number of illuminating studies, al- however, whether they act on origins directly and in a though replication in vitro of a plasmid containing the positive fashion must await further studies. In another origin has yet to be achieved. The origin snup consists exciting development, the cdcl8 of Scbizosac- of at least two dimers of the UL9 protein, which pro- charomycespombe has been shown to suppress muta- duces the open complex in an AT-rich region even in tions of cdclO, a transcriptional activator that is needed the absence of ATP [37]. UL9 has helicase activity, but for entry into S phase [46"']. The cdcl8 gene may be a it does not act catalytically like the other dis- major target of cdclO and, consistent with this, cdcl8 cussed thus far, Instead, it must bind DNA stoichiomet- mutants fail to enter S phase [46"']. Further, cdcl8 may rically for unwinding and requires a stretch of ssDNA play a role in checkpoint control, as cdcl8 mutants are for duplex-unwinding activity [38,39]. The UL9 helicase unable to prevent mitosis and rapid cell division even is stimulated by the virally encoded SSB (called ICP8) though the chromosomes are not fully duplicated. The and it seems likely they function together at the origin cdcl8 gene sequence reveals a nucleotide-binding site, to enlarge the open complex for the future replication but whether the encoded protein plays a direct role fork. (e.g. helicase) or an indirect role in replication must await future isolation and characterization of the gene product. Yeast chromosomal origin In the past two years, rapid and exciting advances have been made in the study of the biochemistry of yeast cellular replication. Yeast origins are known as au- Events after helicase entry tonomous replication sequences (ARS) for their ability to confer autonomous replication on plasmids. Recent Once the helicase has entered the chromosome, it pre- studies [40"°] of the ARS1 origin show that it contains sumably nucleates assembly of replication forks con- four sequence elements: the A element, which contains mining the primase and two replicative , an 11 bp ARS core consensus sequence (ACS) common one for each strand of DNA [1]. The leading and lag- to all ARS elements, the B3 element, which is the bind- ging strand polymerases in eukaryotes appear to be ing site of the ABF1 transcription factor, and the B1 and different polymerases (polymerases 8 and e), whereas B2 elements, which may interact with (as yet) uniden- prokaryotes use two copies of the same polymerase. tiffed origin accessory proteins. A large origin recogni- The mechanism of replicase function is the second tion complex (ORC), which binds to ARS1 at the A site, topic of this essay. has now been purified from Saccharomyces cerevisiae [41"], and genomic footprinting experiments indicate the presence of the ORC on ARS1 in vivo [42"']. The ORC contains six subunits and requires ATP to bind the ARS. The exact function of the ORC is not certain, Chromosomal replicases but a number of observations confirm that it plays a Replicases of E. coli, phage T4, yeast and humans central role in replication. First, the ORC does not bind Organisms that span the evolutionary spectrum have to single-site mutant forms of ACS that lack ARS activity been shown to possess replicases that appear to be [41"']. Second, mutations affecting the 70 kDa subunit similar in function and also in their actual structure. of the ORC, identified because they produce a defect in These are the E. coli DNA polymerase III holoen- function of the HMR silencer, show that the gene is es- zyme (Pol III holoenzyme), the phage T4 replicase, sential for cell viability, and a conditional lethal allele is and the polymerase 15 (pol 8) of both yeast and hu- defective in chromosome replication [43"]. Third, each mans. In each of these systems, the replicase encom- of the six encoding the ORC subunits are essen- passes several proteins that use ATP to initiate rapid tial to yeast (SP Bell, R Kobayashi, B Stillman, abstract and highly processive DNA synthesis. It can be thought 15, Eukaryotic DNA Replication Meeting, Cold Spring of as having three components (see Table 1): the Harbor, September 1993). catalytic component, which contains the DNA poly- Recent developments in the study of the genetics of merase and proofreading 3"-5" exonuclease activities, yeast replication have given researchers clues as to the and the two categories of polymerase accessory pro- roles of important replication proteins. Several genes of teins, a complex of accessory proteins and a single S. cerevisiae have been isolated, the products of which subunit factor. The accessory proteins are are needed for minichromosome maintenance (MCM) needed to confer high processivity onto the catalytic of ARS-containing plasmids. Three of these, MCM2, polymerase. Over the past two years, a deeper under- MCM3 and MCM5, are homologous in sequence and standing of the action of these accessory proteins has each is essential for cell viability [44"]. The intracellu- been developed and appears to apply generally to all lar location of MCM2, 3 and 5 is cell cycle regulated, four systems. Studies have shown that the single sub- they are moved into the nucleus upon completion of S unit accessory protein is a DNA-sliding clamp, which phase and then moved into the cytoplasm at the start of tethers the polymerase to DNA for high processivity. DNA replication: enzymology and mechanisms Kelman and O'Donnell 189

The accessory protein complex functions as a 'clamp shows that it is indeed in the shape of a ring capable loader' that couples ATP to assemble the clamp protein of completely surrounding duplex DNA and reveals the on DNA in a two-step assembly process (see Fig. 2). unexpected feature that [3, although only a dimer, has a The clamp loader recognizes the primed template and sixfold appearance (Fig. 3) [49*']. This symmetry is the couples ATP to assemble the sliding clamp onto DNA result of the three globular domains that comprise each to form a pre-initiation complex. This is followed by monomer and the polypeptide chain backbone struc- association of the polymerase to form the initiation tures of these domains are nearly superimposable. The complex (reviewed in [47]). [3 dimer cannot assemble onto DNA independently, The best studied system, at least for this assembly reac- but in fact requires the five-protein y complex clamp tion, is the Pol IIl holoenzyme of E. colL The proces- loader, a molecular matchmaker that hydrolyzes ATP sivity factor of Pol III, the [3 subunit, is a dimer that to assemble the [3 dimer around DNA. The catalytic freely slides along duplex DNA and is topologically component of the Pol Ill holoenzyme, termed Pol III linked to the DNA, in as much as it binds tighdy to a core, assembles with the [3 ring, which tethers the Pol nicked circular plasmid, but upon linearization freely III core to the DNA and continues to slide with it for slides off over the ends [48]. These results have been rapid and highly processive synthesis. The 7 complex explained by the hypothesis that [3 encircles DNA like acts catalytically to assemble multiple ~ clamp pre-ini- a doughnut [48]. The crystal structure of the [3 dimer tiation complexes on different primed templates. The

Table 1. Three part structure of chromosomal replicases of E. co~i, phage T4, yeast and human.

Component E. colt Phage T4 Eukaryotic

I. Polymerase,3'-5' exonuclease Pol III core gp43 pol 8 a (pol) p125 ¢ (exo) p50 B pol¢

II. Accessorycomplex (clamp loader) y-complex gp44-gp62 complex RF-C (A~ivato~l) DNA-dependent 7 gp44 p12B ATPase 8' p37 Binds clamp 8 gp62 p40 Binds SSB X Unknown function v2 p38, p36

III. Processivity factor (sliding clamp) 13 81>45 PCNA

Replicase subunits and subassemblies are isolated as three pieces: the polymerase/exonuclease, the accessory complex (clamp loader), and the processivity factor (sliding clamp). For complexes, the individual subunits are listed along with the present knowledge of its associated function. The E. coli Pol III holoenzyme contains one further subunit called ~ that binds one 7 complex and two Pol III cores. The components of the yeast and human replicases have the same names and very similar structure; therefore, the yeast and human components are grouped under the heading 'Eukaryotic', but the subunit masses listed are particular to the human replicase.

Fig. 2. Assembly of a processive chromo- Pol !11 core somal replicase. In the E. colt system, (polymerase) the 7 complex clamp loader (accessory complex) recognizes the ssDNA-dsDNA junction, binds the [3 subunit sliding clamp (processivity factor), and then cou- ples hydrolysis of ATP to assemble the [3 subuni,~e4....,~AT P ADP, Pi ring-shaped clamp around DNA forming the pre-initiation complex. The 7 complex dissociates from the pre-initiation clamp (slidingclanlp)~s'~3'DNA7 complex ~ "i~-~) and is capable of assembling other I], clamps on other DNA molecules. The Iniliation Pol III core associates with the 13 slid- complex (clamp loader) Pre-initiation ing clamp to form the initiation complex, complex which is capable of highly processive polymerization. This mechanism may ap- ©I 994 Current Opinion in (,enelit • and l~.,vc, lol)n~.-nl ply generally to other replicases. 190 Chromosomes and expression mechanisms

7 complex can be removed from ttle reaction prior do they really form rings like the E. coli ~ subunit? to adding Pol III core without effect on the rate and None of these proteins has a significant level of se- processivity of DNA synthesis (for recent reviews, see quence homology with any other, and a major dif- [47,50,511. ference is that PCNA and gp45 are trimers, whereas 13 is a dimer, and they are only 2/3 the size of 13. How similar are the 3"4 replicase and eukaryotic The sixfold synmaetry of the 13 dimer held the expla- pol 8 to E. coli Pol III holoenzyme? Both "I"4 and nation for these differences in size and aggregation eukaryotes have an accessory complex that is analo- state. A trinaer of PCNA (and gp45) is of similar mass gous to the E. coli clamp loader T complex. In yeast to a dinaer of I~ so, if the PCNA and gp45 monomers and human, the accessory complex is replication factor were two-domain proteins, instead of three, then the C (RF-C, also called Activator-l) (reviewed in [52]) and trimer would contain a total of six domains like the in "I"4, it is the complex of the gene 44 protein (gp44) 13 dimer. Indeed, a sequence alignment of gp45 and and gene 62 protein (gp62) (gp44--gp62, reviewed in PCNA with the first two domains of 1~, using the struc- [53]). These accessory complexes are DNA-dependent ture of 13 as a guide, shows that the hydrophobic core ATPases that are stimulated by their respective clamp residues are positionally conserved, lending strength protein, and in all cases a primed template is the best to this hypothesis 149"']. As of September 1993, the DNA effector. Since 1992, the genes encoding the five answer was determined for the yeast PCNA, and the subunits of human RF-C have been identified, as have answer was a most definite yes, as the X-ray structure the remaining four genes of the five-subunit E. coli T shows it to be a trinaer in the shape of a ring with the complex [54"-57",58--611. Amino acid sequence com- same outside and inside diameters as the 13 dimer parison shows that several subunits of RF-C are homo- (J Kuriyan, personal communication). Further, the logous to the 7 and 8' subunits of E. coli ~l complex affinity of human PCNA to DNA (placed there by RF- and to the T4 gp44 [56"]. The level of homology is C) depends on the geometry of the DNA, being tightly sufficient to predict that they will have similar three- retained on circular DNA, but freely sliding off upon dimensional structures. Despite these structural simi- linearization of the DNA; therefore, PCNA exhibits the larities, it remains to be established whether the T4 same hallmarks of topological binding to DNA as the and eukaryotic accessory complexes are truly clamp E. colf ~ subunit (N Yao, Z Kelman, Z Dong, Z-Q Pan, loaders and whether they need to remain associated J Hurwitz, M O'Donnell, unpublished data). with the polymerase during elongation. Recent experi- ments, however, are consistent with their being clamp Now, back to the issue of whether the accessory pro- loaders and indicate that they may not be needed dur- teins of "I"4 and eukaryotes are clamp loaders, and ing elongation. whether they must be present with the polymerase and clamp protein during elongation. Three recent reports In the "1"4 system, the processivity factor is the gene 45 in these systems indicate that the clamp--polymerase protein (gp45), and in yeast and humans, it is the pro- unit is all that is needed for processivity. Clever ex- liferating cell nuclear antigen (PCNA) (Table 1). But, periments in the yeast system show that on linear

"A .... B"

Dimer ._~ Interface .~ 3

Fig. 3. Crystal structure of the 13 DNA-sliding clamp of DNA polymerase III holoenzyme. The structure on the left is viewed 'face on' looking through the central cavity. The outside is a continuous layer of antiparallel sheet, which also forms the dimer interfaces (arrows). The central cavity is lined with 12 o. helices, the only helices in the entire molecule. The structure on the right is the 13 dimer turned on its side, with duplex DNA modeled through the center. The two structurally distinct faces (A and B) are the result of the head-to-tail arrangement of the subunits. Reproduced from Molecular Biology of the Cell 1992, 3:955 by copyright permission of the American Society for Cell Biology. DNA replication: enzymology and mechanisms Kelman and O'Donnell 191

DNA (but not circular) PCNA confers processivity onto 7,.), but it is not known whether the interaction is medi- pol 8 in the complete absence of ATP and RF-C [62"]. ated by gp44 or gp62 [71]. Using footprinting methods, This result was interpreted as showing that the PCNA laser-induced UV-crosslinking of protein to DNA, and ring threads itself onto the end of linear DNA and then novel DNA-protein cross-linking agents, it has been couples with pol 8 for processive synthesis, thus cir- found that the 1"4 gp44-gp62 complex needs only to cumventing the need to open the ring, an action bind ATP to bring the gp45 clamp to DNA, although that requires the RF-C complex and ATP. In the T4 it is not known whether the clamp surrounds DNA system, evidence that the gp44-gp62 complex is not at this point [72,73"',74]. Further, these studies show needed during elongation has come from the obser- that the gp44-gp62 complex works on the primed vation that supply of a large excess of gp45 increases template junction to place the gp45 clamp on DNA, the processivity of the polymerase in the absence of and then ATP hydrolysis induces a movement or the the gp44-gp62 complex, a result that is similar to dissociation of gp44-gp62 from the DNA, presumably those from previous studies in the E. coli system and to make room for the polymerase [74,75]. which enforces the idea of the clamp-polymerase as the processive unit [63"]. Furthermore, an elegant elec- Much less is known about the RF-C subunits, although tron microscopy study has shown that the T4 sliding with the recent identification of the genes encoding clamp on DNA appears as a 'hash-mark' (with simi- them we can expect much more information in the lar dimensions to the ~ ring) through which DNA is near future. The human p128 is known to bind DNA, threaded, and these 'hash-marks' appear in clusters as it was identified in 1993 in a south-western analy- indicating that they can slide [64"']. The size of the sis as a DNA-binding protein called PO-GA [57"]. At 'hash-mark' is insufficient to acconmlodate the mass the same time, a similar screen resulted in the isola- of both the gp44-gp62 complex and the gp45 trimer tion of the analogous RF-C subunit of the mouse [76"]. and in light of similarity to other systems it is probably The p37 and p40 proteins are the only RF-C subunits a gp45 clamp. to be obtained in pure form thus far and studies have shown that p37 binds DNA, suggesting that p128 and The actual mechanism by which the clamp loader as- p37 may be similar to ~/and 8" (and T4 gp44) [77"]. The sembles the clamp around DNA is a fascinating issue p40 binds ATP and PCNA, which suggests a functional and is being addressed in all these systems, but detailed analogy to 8 (and T4 gp62) [77"]. A summary of these mechanisms are still unknown. Most of the work has individual subunit functions is included in Table 1. been in the T4 and E. coli systems, in which the indi- vidual subunits of the clamp loaders are available in In all these systems, the three replicase components pure form. No individual subunit of either of these have only weak interactions with one another, at least clamp loaders can assemble their respective clamp in the absence of DNA. In the E. coli system, however, onto DNA; presumably this reaction is too compli- there is another subunit called ~ that firmly binds one cated for just one protein. The five subunits of the ~/complex molecule and two molecules of Pol III core, E. coli ~l complex are % 8, 8", X and "q. Although only T thereby acting as ~i scaffold to all of these pro- and 8 are essential to place ~ onto DNA, the 8' subunit teins tightly together in a 'holoenzyme' particle [67]. A stimulates this reaction considerably [65,66]. DNA-de- dimeric polymerase complete with one clamp loader pendent ATPase activity is produced by a mixture of within a single molecular structure fits nicely with the T and 8' [65] and T is the presumed site of hydrolysis, need to synthesize two strands of DNA, one of which as it is known to bind ATP [1]. The 8 subunit forms a is synthesized discontinuously and requires multiple protein-protein complex with ~ [67]. Hence, it appears clamps to be loaded onto it (lagging strand). At present, that ~' recognizes the primed template, and the 8 sub- the T4 and eukaryotic replicases have been purified as unit functions to bring ~ into the structure for assembly the three separable components and whether, in these around DNA and hydrolysis of ATP. At elevated ionic systems, the clamp loaders maintain an association strength, such as exists in the cell, the X and ~t subunits with the polymerase and clamp or have a functional of the T complex are also needed to initiate processive equivalent of "t to hold them together remains to be synthesis [68]. This is probably rooted in the fact that determined. the T complex associates with SSB-coated DNA [69"], an interaction mediated by X that may give the T com- plex the added grip that it needs to bind the template Replicases of HSV-1 and phage T7 in elevated salt (Z Kelman, M O'Donnell, unpublished The other two replicases that are highly processive are data). those of HSV-1 and phage T7 [1]. These replicases re- The T4 clamp loader is composed of a tetramer of gp44 quire only one accessory protein for high processivity, tighdy associated with one protomer of gp62 [53]. By UL42 for HSV-1 and E. colithioredoxin for T7, and they itself, the gp44 is a DNA-dependent ATPase (like E. coli do not need ATP. Thus, they lack a clamp loader. The "#'), implying that it harbors the DNA-binding and crystal structure of thioredoxin bears no resemblance ATPase sites [70]. The DNA-dependent ATPase of the to a ring, so the molecular basis by which the acces- gp44-gp62 complex is stimulated by gp45, whereas the sory protein provides processivity to the polymerase is gp44 ATPase is not, implying that gp62 interacts with not clear. One simple possibility is that the polymerase gp45 (like E. coli 8) [70]. The gp44-gp62 complex in- has a cleft into which the DNA fits and the accessory teracts with gp32 (T4 SSB) bound to ssDNA (like E. coli protein seals off this cleft to trap DNA inside. 192 Chromosomes and expression mechanisms

Interaction of sliding clamps with. other proteins tems is revealing fundamental diversification. The new information about yeast origin structure, and the pro- The clamp proteins have recently been shown to share teins that act on it, lead us to anticipate future discov- a common property, in that they interact with pro- eries concerning the mechanisms by which eukaryotic teins other than the replicative polymerase. The best cellular origins are activated, regulated and eventually studied case is that of T4, in which the gp45 clamp integrated with signal transduction. interacts with RNA polymerase. Phage T4, as well as several viruses (e.g. SV40, adenovirus and HSV), has an early-to-late switch in gene expression, so that late genes (eg capsid proteins) are not expressed until the References and recommended reading viral genome has been replicated. Elegant studies by Peter Geiduschek's research group have revealed that Papers of particular interest, published within the annual perkxt of the underlying mechanism of the early-to-late switch review, have been highlighted as: is an interaction between the T4 replicase accessory • of special interest proteins and the E. coli RNA polymerase (modified by • . of outstanding interest "1"4 gp33 and gp55). In 1992, a continuation of these 1. KORNBERGA, BAKERTA: DNA Replication. New York: WH studies was published showing that this switch is the Freeman; 1991. result of a tracking mechanism, whereby the accessory 2. ECHOLS H: Multiple DNA-Protein Interactions Govern- proteins assemble at a nick and slide along DNA, thus ing High-Precision DNA Transactions. Science 1986, acting as a 'mobile enhancer' for late gene activation 233:1050--1056. [78"']. Presumably the interaction is through the gp45 3. HWANGDS, CROOKE E, KORNBERG A: Aggregated DnaA clamp and the modified RNA polymerase, although a Protein is Dissociated and Activated for DNA Replication by Phospholipase or DnaK Protein. J BIol Chem 1990, role for the gp44-gp62 complex cannot be ruled out, 265:19244-19248. as it was not removed from the system after formation of the sliding clamp. This mechanism may apply to 4. HWANGDS, KORNBERG A: Opening of the Replication Origin • . of Escberlcbla coil by DnaA Protein With Protein HU or eukaryotic viruses in general, as a baculovirus protein IHE J Btol Cbem 1992, 267:23083-23086. with 42% identity to PCNA is also needed for activation Important demonstration that either protein HU or IHF will suffice of late genes [79]. to form the open complex on negatively supercoiled plasmid filled with DnaA protein complexed with ATP. Use of the sliding clamp for other DNA metabolic pro- 5. SCHMIDMB: More Than Just 'Histone-Like' Proteins. Cell cesses appears to be the rule rather than the exception, 1990, 63:451--453. as the clamps of E. coli and of human also interact with 6. SKARSTADK, THONY 13, HWANG DS, KORNBERG A: A NO- other proteins. In E. coli, the ~ clamp is utilized by Pol ** vel Binding Protein of the Origin of the Escherfcbla coil III and DNA polymerase II, and in humans and yeast, Chromosome. J Btol Cbem 1993, 268:5365-5370. the PCNA clamp is utilized by both pol 8 and pol Identification of a novel origin-binding protein in E. coil The Rob [1]. In fact, the human PCNA has recently been shown protein is present at 5000 copies per cell, contains a helix-turn-helix to bind to members of the cyclin D and cyclin-depen- motif and binds at the right border of ortC. dent kinase families [80"']. The function of the latter 7. HWANGDS, KORNBERG A: Opposed Actions of Regulatory interaction is unknown. Would the PCNA ring need to • * Proteins, DnaA and IciA, in Opening of the Replication Ori- be on or off DNA to manifest the biological activity? It gin of E$cherlchla coll. J Blol Chem 1992, 267:23087-23091. First demonstration that DnaA protein interacts with the middle and is tempting to speculate that the PCNA ring may tether right 13-mers of orICwith sequence specificity, and that lciA protein cyclins and their associated kinases to DNA for some blocks this interaction by binding all three 13-mers. aspect of cell cycle control (e.g. mitotic checkpoint). 8. HWANGDS, THONY B, KORNBERG A: IdA Protein, a Spe- • - cific Inhibitor of Initiation of Escberlcbla coil Chromosomal Replication. J Btol Cbem 1992, 267:2209-2213. Elevation of the intracellular concentration of IciA protein produces a lag in cell growth upon transfer to fresh medium. This correlates Conclusions with a fourfold increase of IciA protein in the stationary phase.

9. CROOKEEC, CASTUMACE, KORNBERG A: The Chromosome In summary, knowledge is expanding rapidly about • - Origin of Escberlcbla coil Stabilizes DnaA Protein Dur- replication mechanisms in several systems, too many, ing Rejuvenation by Phospholipids. J Bfol Cbem 1992, in fact, for them all to have been mentioned here. 267:16779-16782. Chromosomal replicases appear to be quite similar in First demonstration that not all of the 20 or more DnaA molecules bound to orfC need to be in the ATP-bound form for origin activa- their mechanism for attaining high processivity. How- tion; some can remain complexed with ADP. ever, litde is known about how they assemble at the 10. JACOB F, BRENNER S, CtJZBN F: On the Regulation of DNA origin or communicate with the helicase and primase, Replication in Bacteria. Cold Sprtng Harb Syrup Quant BIol or how these processive enzymes release DNA rapidly 1963, 28:329-348. and rebind new primers during discontinuous synthe- 11. BAKERT, KORNBERG A: Transcriptional Activation of Initia- sis of the lagging strand. Likewise, the outline of events tion of Replication from the E. coil Chromosomal Origin: in activation of origins has been conserved in evolution an RNA-DNA Hybrid Near orlC. Cell 1988, 55:113-123. from prokaryotes to eukaryotes, but important details 12. ALLENGC, KORNBERG A: Fine Balance in the Regulation in the manipulation of these sequences by proteins, the of DnaB Helicase by DnaC Protein in Replication in Es- roles of ATP, and regulation of initiation in different sys- cherlchla coll. J Btol Cbem 1991, 266:22096-22101. DNA replication: enzymology and mechanisms Kelman and O'Donnell 193

13. SCHNOS M, ZAHN K, INMAN RB, BI.A'FI'NER FR: Initiation 29. CHENGL, WORKMAN JL, KINGSTON RE, KEI.LY mJ: Regulation Protein Induced Helix Destabilization at the ~. Origin: a ,,. of DNA Replication In Vitro by the Transcriptional Activa- Prepriming Step in DNA Replication. Cell 1988, 52:385-395. tion Domain of GAL4-VP16. Proc Nail Acad Scl USA 1992, 89: 589--593. 14. ALFANOC, MCMACKEN R: Ordered Assembly of Nucleopro- It is hypothesized that transcription fhctors either interfere with as- rein Structures at the Bacteriophage ~. Replication Origin sembly of chromatin near the origin or facilitate binding of initiation During the Initiation of DNA Replication. J Blol Chem 1989, factors at the origin. Whatever the mechanism, it may be similar to 264:10699-10708. the role these factors play in transcription activation. 15. MALLOTYJB, ALFANO C, MCMACKEN R: Host Virus Interac- tions in the Initiation of Bacteriophage ~. DNA Replication. 30. VIRSHUPDM, RUS.¢,OAAR, KELLYTJ: Mechanism of Activation Recruitment of Escherichla coil DnaB Helicase by ~. P Repli- ss of Siman Virus 40 DNA Replication by Protein Phosphatase cation Protein. J Btol Chem 1990, 265:13297-13307. 2A. Mol Cell Btol 1992, 12:4883--4895. The two hexamers of T-antigen assemble on the origin in two dis- 16. ALFANOC, McMACKEN R: Heat Shock Protein-Mediated Dis. tinct stages. This paper demonstrates that this is regulated by pro- assembly of Nucleoprotein Structures Required for the Ini- tein phosphatase 2A. The major influence of protein phosphatase tiation of Bacteriophage ~. DNA Replication. J Btol Chem 2A treatment is that it enhances co-operative binding of the .second 1989, 264:10709-10718. hexamer to the origin. 17. LIBEREKK, GEORGOPOULOS C, ZYLICZ M: Role of the Es- 31. DLrITAA, ROPPERTJM, A.q'I'ERJC, WINCHF.STER E: Inhibition cherlchla coil DnaK and DngJ Heat Shock Proteins in the ss of DNA Replication Factor RPA by p53. Nature 1993, Initiation of Bacteriophage ~. DNA Replication. Proc Nail 365:79--82. Acad Set USA 1988, 85:6632--6636. The authors show that p53 can bind both T-antigen and RP-A at 18. MENSA-WILMOTK, CARROLL K, MCMACKEN R: Transcriptional the .same time and, consistent with this, a different region of p53 Activation of Bacteriophage ~. DNA Replication In Vitro: is needed to bind replication protein A than to bind T-antigen. Regulatory Role of Histone-Like Protein HU of Escberichla 32. DtfITAA, S'nLLMAN B: cdc2 Family Kinases Phosphorylate coll. EMBO J 1989, 8:2393-2402, • . a Human Cell DNA Replication Factor, RPA, and Activate 19. WICKNER S, SKOWYRA D, HOSKINS J, MCKENNEY K: DnaJ, DNA Replication. EMBO J 1992, 11:2189-2199. ss DnaK, and GrpE Heat Shock Proteins are Required in orlP1 Stimulation of SV40 replication by cdc2 kinase was performed in DNA Replication Solely at the RepA Monomerization Step. cell extracts. The origin unwinding reaction was "also stimulated. An Proc Nail Acad Sol USA 1992, 89:10345--10349. unidentified factor in the extract is required for stimulation. Urea treatment monomerizes RepA and completely bypasses the re- quirement for heat shock proteins for in t~tro replication of an oriP1 33. SEO Y-S, MULLER F, LUSKY M, HORWITZ J: Bovine Papilloma plasmid, showing that heat shock proteins are required only at the s. Virus (BPV)-Encoded E1 Protein Contains Multiple Activities monomerization step. Required for BPV DNA Replication. Proc Nail Acad Set USA 1993, 90:702-706. 20. BAKERTA, WICKNER SH: Genetics and Enzymology of DNA This paper shows that the E1 helicase is most efficient on a forked Replication in Escherlchla coll. Annu Rex, Genet 1992, DNA template and is supported by all eight nucleoside triphospates. 26:447-477. 34. WANG L, MOHR I, Fom.x E, LIM DA, NOHAILE M, BOTCHAN 21. CHALLH,ERG MD, KELLY TJ: Animal Virus DNA Replication. • * M: The E1 Protein of Bovine Papilloma Virus 1 is an ATP- Annu Rev Blochem 1989, 58:671-717. Dependent DNA Helicase. Proc Naa Acad Sol USA 1993, 22. MCVEY D, BRIZUELA L, MOHR l, MAILSHAK DR, GLUZMAN 90:5086-5090. Y, BEACH D: Phosphorylation of Large Tumor Antigen Important demonstration that at high levels of El, and in the absence by cdc2 Stimulates SV40 DNA Replication. Nature 1989, of E2, plasmids lacking the origin are unwound and replicated. This 341:503-507. paper discusses how this promiscuous origin-independent replica- tion relates to earlier observations in some eukaryotic systems (e.g. 23. SCHEII)MANN KH, VIRSHUP DM, KELLY TJ: Protein Phos- Xenopus), which implied that cellular chromosomes do not have de- phatase 2A Dephosphorylates Siman Virus 40 Large T fined origins. Antigen Specifically at Residues Involved in Regulation of DNA-Binding Activity. J Virol 1991, 65:2098-2101. 35. LUSKYM, HURWrlXJ, SEO Y-S: Cooperative Assembly of the • Bovine Papilloma Virus El and E2 Proteins on the Replica- 24. WANG EH, FRIEDMAN PN, PRIVE.'; C: The Murine p53 Protein tion Origin Requires an Intact E2 Binding Site. J Blol Chem Blocks Replication of SV40 DNA in Vitro by Inhibiting the 1993, 268:15795-15803. Initiation Functions of SV40 Large T Antigen. Cell 1989, The authors demonstrate that proper spacing between the El and E2 57:379-392. sites is required for cooperative binding of these proteins to DNA. 25. lk.)ROWIECJA, DEAN Fig, BULLOCK PA, HURWI'I'Z J: Binding and Unwinding -- How T Antigen Engages the SV40 Origin 36. WANG L, MOHR l, LI R, NOTI'OLI T, SUN S, BOTCHAN M: of DNA Replication. Cell 1990, 60:181-184. Transcription Factor E2 Regulates BPV-1 DNA Replication in Vitro by Direct Protein-Protein Interaction. Cold Spring 26. DEAN FB, BOROWIEC JA, EKI T, HURWITZ J: The Simian Harb Syrup Quant Biol 1991, 56:335-346. s. Virus 40 T Antigen Double Hexamer Assembles Around the DNA at the Replication Origin. J Blol Chem 1992, 37. KOFFA, SCHWEDI-2SJF, TFGTMEYER P: Herpes Simplex Virus 267:14129-14137. Origin-Binding Protein (UL9) Loops and Distorts the Viral In the absence of DNA, ATP promotes hexamerization of T-antigen. Replication Origin. J Vtrol 1991, 65:3284-3292. This hexamer is stable to isolation on a glycerol gradient, but cannot 38. BRUCKNERRC, CRUTEJJ, DODSON MS, LEHMAN IR: The Herpes bind the origin. Incubation without ATP monomerizes T-antigen and Simplex Virus 1 Origin Binding Protein: a DNA Helicase. J activates it for proper assembly onto the origin upon adding ATP and DNA. This suggests that the hexamers assemble to encircle the origin Blol Chem 1991, 266:2669-2674. DNA. 39. FtERERD, CHALLBERG MD: Purification and Characterization 27. MAS'I'RANGELO1A, HOUGH PVC, WALL JS, DOt)SON M, DEAN of ULg, the Herpes Simplex Virus Type 1 Origin-Binding FB, HURWI'IX J: ATP-Dependent Assembly of Double Hex- Protein. J Virol 1992, 66:3986-3995. amer of SV40 T Antigen at the Viral Origin of DNA 40. MARAHRENSY, STILLMAN 13: A Yeast Chromosomal . Nature 1989, 338: 658-662. ss DNA Replication Defined by Multiple Functional Elements. 28. CHENGL, KELLY TJ: Transcriptional Activator Nuclear Factor Science 1992, 255:817-823. I Stimulates the Replication of SV40 Minichromosomes In An important study in which thirty-four overlapping linker insertion Vlvo and in Vitro. Cell 198% 59:541-551. mutants were used to define the precise boundaries of the four ele- 194 Chromosomes and expression mechanisms

ments of ARSl. The multi-element nature of 0J~S1 was correlated to ~i. CHENM, PAN Z-Q, HUR~aTZ J: Studies of the Cloned 37-kDa the known complexity of transcriptional promoters. • Subunit of Activator 1 (Replication Factor C) of Hela Cells. Proc Nail Acad Sol USA 1992, 89:5211-5215. 41. BELL SP, SllLLMAN B: ATP-Dependent Recognition of Eukary- The sequence of p37 of RF-C has a high degree of homology to that • . otic Origins of DNA Replication by a Multiprotein Complex. of p40 of RF-C. Nature 1992, 357:128-!34. This paper describes the purification of the origin recognition com- 55. CHEN M, PAN Z-Q, HURWI'I-LJ: Sequence and Expression plex (ORC) through preparation of yeast nuclear extract. DNA-bind- • in Escberlchla coil of the 40-kDa Subunit of Activator 1 ing activity was detectable only after ion exchange cl~'omatography, (Replication Factor C) of Hela Cells. Pro(: Naa Acad Sol and even then ATP was needed. USA 1992, 89:2516-2520. This paper reports limited .sequence homology between the p40 sub- 42. DIm.EYJFX, COCKERJH: Protein-DNA Interactions at a Yeast unit of RF-C, "1"4 gp44 and E. cull ~/:rod 3. • * Replication Origin. Nature 1992, 357:169-172. The genomic footprint of ARS1 indicates that ORC is situated on the 56. O'DONNELLM, ONRUS'I" R, DEAN FB, CHEN M, HURWI'IZ J: A element and ABF1 on B3. An additional footprint over element B2 • Homology in Accessory Proteins of Replicativ¢ Polymerases may indicate that another protein is present. The results also suggest E. coil to Humans. Nucleic Acids Re• 1993, 21:1-3. that ORC may remain bound to AKS1 throughout the cell cycle. The amino acid sequences of E. coital, "¢ and 8', T4 gp44 and human RF-C subunits 1340, p38, 1337 and p36 are all homologous, implying 43. MICKLEMG, ROWLEY A, HARWOOD J, NA.SM~'I'H K, DIFFLEY a common ancestor. • JFX: Yeast Origin Recognition Complex is Involved in DNA Replication and Transcriptional Silencing. Natttre 1993, 57. Lu Y, ZEVl" AS, RIEGEL T: Cloning and Expression of a Novel 366:87-89. • Human DNA Binding Protein, PO.GA. Blocbem Btophys Re• This study provided the first genetic evidence that ORC is needed Cumin 1993, 193: 779-786. for replication. The PO-GA protein (p120 of RF-C) sequence has homology to E. colt and yeast DNA ligases. 44. YAN H, MERCHANTAM, TYE BK: Cell Cycle-Regulated Nuclear • Localization of MCM2 and MCM3, which are Required for 58. CAR'IER JR, FRANDEN MA, AEBERSOLD R, MCHENRY CS: the Initiation of DNA Synthesis at Chromosomal Replication Molecular Cloning, Sequencing, and Overexpression of the Origins in Yeast. Genes Dev 1993, 7:2149-2160. Structural Gene Encoding the 8 Subunit of Escherlchla This report describes studies showing that defective MCM2 and coil DNA Polymerase Ill Holoenzyme. J Bacterlol 1992, MCM3 proteins result in defects in origin-specific replication. 174:7013-7025. 45. CHENY, HENNESSY KM, BO'I~'rEtN D, TYE B-K: CDC46/MCM5, 59. CARTER JR, FRANDEN MA, AI-'BERSOLD R, MCHENRY CS: • - a Yeast Protein whose Subcellular Locafization is Cell Cycle- Identification, Isolation, and Characterization of the Struo Regulated, is Involved in DNA Replication at Autonomously rural Gene Encoding the 8' Subunit of Escherichla Repficating Sequences, Proc Natl Acad Set USA 1992, coil DNA Polymerase Ill Holoenzyme. J Bacteriol 1993, 89:10459-10463. 175:3812-3822. MCM5 is shown to be identical to CDC46, and mutants are shown to 60. DUNGZ, ONRUST R, SKANGALISM, O'DONNELL M: DNA Poly- be defective in maintaining minichromosomes. The paper describes merase III Accessory Proteins. i. holA and holB Encoding the cell cycle regulated subcellular localization of MCM5. 8 and 5'. J BIol Cbem 1993, 268:11758-11765. 46. KELLYTJ, MARTIN GS, FORSBURG SL, S'IEPHEN RJ, RUSSO A, 61. XIAO H, CROMBIE R, DONG Z, ONRUSW R, O'DONNELL M: • * NURSE P: The Fission Yeast cdcl8 • Gene Product Couples DNA Polymerase Iii Accessory Proteins. 111. holC and hold S Phase to START and Mitosis. Cell 1993, 74:371-382. Encoding X and ~. J Btol Chem 1993, 268:11758--11765. Checkpoint control is a process that prevents mitosis until after the completion ors phase. Thus, the cdc18 product functions at initiation 62. BURGERSPM, YODER BL: ATP-Independent Loading of the and termination of S phase. • * Proliferating Cell Nuclear Antigen Requires DNA Ends. J BIol Chem 1993, 268:19923-19926. 47. O'DONNELL M, KURIYAN J, KONG X-P, STUKENBERG PT, PCNA was shown to be capable of loading omo the end of a long ONRUST R: The Sliding Clamp of DNA Polymerase llI Hoioen- stretch of duplex DNA (0.55 kb), implying that it can slide long dis- zyme Encircles DNA. Moi BIol Cell 1992, 3:953-957. tances, but it could not load over ssDNA coated with SSB. These are 48. STUKENBERG PT, STUDWEI.L-VAUGHAN PS, O'DONNELL M: similar mobility characteristics to those of the E. c-off ]3 sliding clamp. Mechanism of the ]3-Clamp of DNA Polymerase IIi Holoen- 63. REDDYMK, WEITZEL SE, VON HIPPLE PH: Assembly of a Func- zyme. J Biol Chem 1991, 266:11328-11334. • tional Replication Complex without ATP Hydrolysis: a Direct 49. KONGX-P, ONRUST R, O'DONNELL M, KURIYAN J: Three Di- Interaction of Bacteriophage "1"4 gp45 with T4 DNA Poly- • • mensional Structure of the ]3 Subunit of Escherlchla coil merase. Proc Natl Acad Set USA 1993, 90:3211-3215. DNA Polymerase III Holoenzyme: a Sliding DNA Clamp. Cell An interesting observation that use of a large amount of gp45 re- 1992, 69:425-437. sults in clamp assembly on a circular DNA molecule without the This report shows that the [3 subunit is in the shape of a ring, as gp44-gp62 complex or ATP. predicted [48], and the authors present the new hypothesis that pro- 64. GOGOLEP, YOUNG MC, KUBASEKWL, JARVIS TC, VON HIPPEL liferating cell nuclear antigen (PCNA) and gp45 are trimers with a • . PH: Cryoclectron Microscopic Visualization of Functional similar shape to that of ]3, based on amino acid sequence compar- Subassemblies of the Bacteriophage T4 DNA Replication isons and the structure of ~. Complex. J Mol Blol 1992, 224:395--412. 50. O'DONNELL M, KUPJYAN J, KONG X-P, STUKENBERG PT, The observed T4 DNA replication complex structures (hash-marks) ONRUST R, YAO N: The ]3 Sliding Clamp of E. coil DNA were not formed on supercoiled DNA, but only on DNA with a nick Polymerase Ill Balances Opposing Functions. Nucleic Acids or a gap. In some instances several hash-marks were observed in Mol Bfol 1993, 8:in press. clusters on one DNA molecule. It is not yet certain whether the observed hash-mark structure is the gp45 trimer or the gp44--gp62 51. KURIYANJ, O'DONNELL M: Sliding Clamps of DNA Poly- complex. merases. J Mol BIol 1993, 234:915-925. 65. ONRUST R, O'DONNELL M: DNA Polymerase IlI Accessory 52. DOWNEY KM, TAN C-K, SO AG: DNA Polymera~ Delta: Proteins. H. Characterization of 8 and 8". J BIol Chem 1993, a Second Eukaryotic DNA Replicase. Bide•says 1990, 268:11766-11772. 12:231-236. 66. ONRUSTR, STLIKENBERG PT, O'DONNELL M: Analysis of the 53. YOUNGMC, REDDY MK, YON HIPPLE PH: Structure and Func- ATPase Subassembly which Initiates Processive DNA Synthe- tion of the Bacteriophage T4 DNA Polymerase Holoenzyme. sis by DNA Polymerase III Holoenzyme. J BIol Chem 1991, Biochemistry 1992, 31:8675--8690. 266:21(-~1-21686. DNA replication: enzymology and mechanisms Kelman and O'Donnell 195

67. ONRUSTR: The Structure and Function of the Accessory zyme at a Primer-Template Junction. J Biol Chem 1991, Proteins of the E. coil DNA Polymerase Ill Holoenzyme. 266:20034--20044. [PhD Thesis]. New York: Comell University Medical College; 1993. 76. BURBELOPD, UTANI A, PAN Z-Q, YAMADAY: Cloning of the • Large Subunit of Activator 1 (Repfication Factor C) Reveals 68. XIAOH, DUNG Z, O'DONNELL M: DNA Polymerase III Acces- Homology with Bacterial DNA I igases. Proc Nail Acad Scl sory Proteins. VI. Characterization of c and y. J Btol Chem USA 1993, 90:11543--11547. 1993, 268:11779-11784. This paper describes the cloning of the gene encoding RF-C. The au- 69. FRADKINLG, KORNBERG A: Prereplicative Complex of Corn- thors note regions of homology between the carboxy-terminal por- • ponents of DNA Polymerase III Holoenzyme of Escherlchla tion of the large subunit of murine RF-C and the four smaller subunits coil. J BIol Chem 1992, 267:10318-10322. of human RF-C. This study identifies an association between y complex and ssDNA 77. PAN Z-Q, CHEN M. HuRwrrz J: The Subunits of Activator 1 that requires SSB. The paper also shows that the "y complex can place • (Replication Factor C') Carry Out Multiple Functions Essen- [3 onto DNA without SSB. tial for Proliferating-Cell Nuclear Antigen-Dependent DNA 70. RUSH J, LIN T-C, QUINONES M, SPICER EK, DOUGLAS I, Synthesis. Proc Nail Acad Sci USA 1993, 90:6--10. WILLIAMS KR, KONIGSBERG WH: The 44p Subunit of the "1"4 The authors show that under some conditions p37 stimulates pol 8, DNA Polymerase Accessory Protein Complex Catalyzes ATP but not pol 8. They also show that p40 inhibits pol 5, suggesting a Hydrolysis. J Biol Chem 1989, 264:10943-10953. p40-pol 8 interaction. 71. CHA T-A, ALBER'I~ BM: In Vltro Studies of the T4 Bacterio- 78. HERENDEENDR, KASSAVE'nSGA, GEIDUSCHEK EP: A Transcrip- phage DNA Replication System. Cancer Celb 1988, 6:1-10. • . tional Enhancer whose Function Imposes a Require- ment that Proteins Track Along DNA. Science 1992, 72. MUNH MM. ALBER'I.'S BM: The T4 DNA Polymerase 256:1298-1303. Accessory Proteins Form an ATP-Dependent Complex The nicked site for assembly of the T4 sliding clamp was placed on on a Primer-Template Junction. J Blol Chem 1991, one DNA ring of a two ring catenane; the other ring contained the T4 266:20024-20033. late promoter. The sliding clamp on one ring was unable to enhance 73. HOCKENSMrI'HJ'~/, KUBASEK WL, EVER'I~Z EM, MESNER LD, transcription on the other ring. • • VON HIPPEL PH: Laser Cross-Linking of Proteins to Nucleic Acids. II. Interactions of the Bacteriophage T4 DNA Repli- 79. O'REILLYDR, CRAWFORD AM, MILLER LK: Viral Proliferating cation Polymerase Accessory Proteins Complex with DNA. Cell Nuclear Antigen. Nature 1989, 337:606. J B~ol Chem 1993, 268:15721-15730. 80. XIONGY, ZHANG H, BEACH D: D Type Cyclins Associated This paper shows that UV-induced cross-links between DNA and • - with Multiple Protein Kinases and the DNA Replication and gp62 are increased upon adding gp45 to the gp44-gp62 complex, Repair Factor PCNA. Cell 1992, 71:505-514. indicative of a conformational change of gp44--gp62 relative to DNA It is proposed that PCNA is a member of multiprotein complexes upon placing gp45 on DNA. The authors present models by which containing p21 and combinatorial variations of cyclins D1 and 3 and gp45 is placed on DNA by gp44-gp62 complex. cyclin-dependent kinases 2, 4 and 5. 74. CAPSONTL, BENKOVIC SJ, NOSSAL NG: Protein-DNA Cross- Linking Demonstrates Stepwis¢ ATP-Dependent Assembly of T4 DNA Polymerasc and its Accessory Proteins on the Primer-Template. Cell 1991, 65:249-258. Z Kelman and M O'Donnell, Department of Microbiology, Comell 75. MUNN MM, ALBERTS BM: DNA Footprinting Studies of the University Medical College, 1300 York Avenue, New York, New York Complex Formed by the T4 DNA Polymerase Holoen- 10021, USA.