J. k’ol.

Biol. (1991) 222, 127-131

COMMUNICATIONS

Initiation

of Bacteriophage PRDl DNA Replication Single-stranded Templates

on

Seung-Ku Yoo and Junetsu Ito Department of Microbiology and Immunology College of Medicine, The University of Arizona Tucson, AZ 85724, UL3.A. (Received 8 March

1991; accepted 23 July

1991)

In vitro studies have demonstrated that single-stranded DNA molecules containing the 3’ terminal nucleotides of the PRDl DNA replication origin can support initiation by a protein-primed mechanism. We have determined the minimal requirements for priming by analyzing the template activity of various deletion derivatives. Our results showed that the 3’-terminal 15 nucleotides of the replication origin are sufficient for priming. The finding that the requirements for recognition of replication origin are different from those for priming suggests that there are two distinct steps during initiation of PRDl DNA replication: first, recognition of the replication origin on double-stranded DNA and second, the priming event on single-stranded DNA. In addition our results showed that additional bases at the 3’ end of templates did not affect priming activity, suggesting that the priming site is searched for from inside the template. Keywords: linear DNA replication;

origin of replication;

Genome-linked proteins of viruses are wide spread in nature (Wimmer, 1982). These proteins, called terminal proteins, are all bound via a phosphodiester bond to the 5’-terminal nucleotide of the genome, an observation that immediately suggests the involvement of the protein in viral genomic replication, possibly at the stage of initiation (Rekosh et al., 1977). Recent evidence has clearly shown that this hypothesis (protein-priming mechanism) is indeed the case (Challberg & Kelly, 1989). The protein-priming mechanism is regarded as a novel way to preserve the 5’ ends of linear DNA during replication (Challberg & Kelly, 1979; Stillman et al., 1981; Lichy et al., 1981; Watabe & Ito, 1983; Watabe et al., 1984; Blanc0 & Salas, 1984). Three DNA virus groups (adenovirus-2, 429, and PRDl) have been extensively studied to understand the events that occur during protein-primed DNA replication (Challberg & Kelly, 1989; Stillman, 1989; Salas, 1988; Hsieh et al., 1990; Yoo & Ito, 1991). Each DNA replication system has its own unique characteristics, although all three systems have much in common. It is of interest to compare these three virus systems in order to gain greater insight into the molecular basis of the replication process, the ultimate purpose of which is the preservation of the entire genetic information throughout propagation. 0022%2836/91/22012745

$03.00/O

recognition;

priming site

The nucleotide sequences of the minimal replication origins of the viruses mentioned above (Wides et al., 1987; Gutierrez et al., 1988; Yoo & Ito, 1991) are shown in Figure 1. These three origins do not show any sequence homology. Furthermore, each nucleotide sequence contains unique features. The adenovirus origin (Fig. l(a)) contains two short repeat sequences (CAT and ATA). Whereas the PRDl origin (Fig. l(c)) has two complementary sequences (GGGGA, CCCCT and GGGG, CCCC), the 429 origin (Fig. l(b)) shows neither repeat nor complementary sequences. Therefore, the sequence requirements for DNA replication may differ considerably between the systems. For adenovirus DNA replication, the core sequence (9 to 18 bpt) is absolutely required. The first three A residues are important to 429 DNA replication. The complementary sequences seem to be required for efficient PRDl DNA replication. The GC/AT content of the adenovirus origin contrasts with the PRDl origins. In the adenovirus origin 13 out of 18 nucleotides are A. T base-pairs, whereas 15 out of 20 nucleotides in the PRDl origins are G-C! base-pairs. The aspects described above further suggest that the molecular t Abbreviations used: bp, base-pair(s); dGMP, 2’-guanosine 5’-phosphate; ITR, internal terminal repeat(s); -mer, minimum base elongation requirement.

127

0

1991 Academic

Press Limited

128

S.-K.

(a)

5’.CATCATC+TAATATACCI-3’ spacer core

(adenovirus)

(b)

S’AAAGTAGGGTAC-3’

(429)

(c)

VGGGG~CCC~3 I 2 3 d”fWJl”S

Yoo and J. ito

,PRDI) 4

Figure 1. The nucleotide sequences of the minimal replication origins of protein-primed DNA replication yvstems. (a) The spacer and core region of adenovirus DNA replication origin. The core region (9 to 18 bases) is hoxed. (b) The minimal replication origin of 429. (c) The minimal replication origin of PRDl Domains 1 and 3 are hold-faced, and domains 2 and 4 are boxed.

oasis of recognition for DNA replicatjion origins may be different among these protein-primed DNA replication systems. Sacteriophage PRDl is a lipid-containing phage infecting Escherichia coli, Salmonella typhimurium and other gram-negative bacteria that harbor plasmids of the P, N, or W incompatibility types (Olsen et al., 1974; Bradley & Rutherford, 1975). The receptor for PRDl is believed to be plasmid-coded i)ili (Bradley, 1976). The genome of PRDl is a *inear, double-stranded DNA of about 14,700 basepairs (Bamford et aE., 1983). A terminal protein (molecular mass, 29 kDa) is covalently linked to the 3’ end of PRDI DNA. The linkage between the terminal protein and PRDl DNA is a phosphodiester bond between a tyrosine residue and dGMP. the terminal nucleotide of the PRDl genome f Bamford & Xindich, 1984). The PRDI DNA contains perfect ITR sequences of 110 or 111 basepairs (Savilahti & Bamford, 1986; Gerendasy & Tto, 1987) and the terminal 20 base-pairs of the ITR are required for efficient in vitro DNA replication (Yoo & Ito, 1991). We have shown recently that singlestranded templates containing a PRDl replication origin (20 bases) are capable of supporting initiation of DNA replication in vitro (Yoo & Ito, 1991). In order to define the minimal nucleotide sequence requirements of 3’-terminal bases for priming in vitro several oligonucleotides, having various deletions, were synthesized and tested for initiation assay. Our results show that the 3’-terminal 15 rmcleotides of the PRDl DNA replication origin are absolutely required for the priming reaction in vitro.

Initiation

with synthetic oligonucleotides various deletions

containing

Figure 2 shows that template e, which has only the first 15 bases of the PRDl replication origin, supports priming efficiently. We have also confirmed, using [a- 32P]dATP, that the initiation complex PRDl terminal protein-dGMP can be efficiently elongated on template e (data not shown). The templates that have only the first five nucleotides (template c), or ten nucleotides (template d) did not support priming. Thus, the presence of the

3’-terminal 15 nucleotides was absolut,ely required for initiation to occur. Template e (I 5-mer) appears slightly more effective than template f (20-mer) and as effective as template g (27-mer), which is 12 bases longer than template e. These observations suggest that the length of the template has little effect on priming as long as the templates have the first 15 bases of the PRDl replication origin. The oliyuonucleotides with homo-polymer sequences (templates a and b) did not support priming.

Initiation

with synthetic

oligonucleotides

containing

There are two complementary sequences in the minimal origin of replication of PRJ)l_ Point. mutations in these complementary sequence regions markedly reduced the template activity (Yoo $ lto. 1991). The minimal replication origin can be divided into four sequence domains; domain 1 (C(WT), domain 2 (ATGCA), domain 3 (CGGGG). and domain 4 (AGGGG) (Fig. I). Domains 1 and 4 are perfectly complementary but inverted. The four consecutive c’ residues of domain I. and t,he four consecutive G residues of domain 3 are also complementary. Domain 2 is not complementary to any ot the other three domains. To examine the significance of the complementary sequences we utilized synthetic oligonucleotides with mutations in domains 1, 3 and 4 (Table 1). A 20-base oligonu&~otide containing mutated domains 3 and 4 with sequences not complementary t,o that. of domain I (template 2) did not support priming. Mutations m domain 1 (template 3) or domain 3 (template 5) completely abolished the template activity. On t,hrl other hand, an oligonucleotide with mutations in domain 4 (template 6) did support priming. These results are consistent with the fact. that the first 15 bases of the 3’-terminal bases of the PRDl replication origin are required for priming. Domain 4 does not seem to contribute to priming. Template 4, which bears mutations in domains I, 3 and 4. did not support, the initiation reaction even though the mutated domain 1 (CGGGA) is complementary to domains 3 (CCCCG) and 4 (TCCCG). These observations suggest that the intact complementary sequences themselves are perhaps as important’ as the parts of the specific recognition site for priming. To define the template requirement further, six different mutated templates were prepared and tested for priming in vitro. The results (Table I) clearly show that deletions (templates 8, 9 and 10) or insertions (templates II and 12) in domain 2 demolished the priming activity. suggesting t,hwt the distance between domains 1 and 3 was crucial for priming. Only template 10 csontaining thcx internal sequence TGC of domain 2 showed measur able priming activity compared with that’ of wiltltype (template 7). When the internal sequence of domain 2 was substituted with a non-specific sequence AAA (template 13), the priming act.ivity also decreased drastically.

129

Communications

;: : e f

g

Sequence

Location

3'-ccccccccccccccccccccccccccc-5' 3'-GGGGGGGGGGGGGGGGGGGGGGGGGGG-5' 3'-CCCCT-5' 3’-CCCCTATGCA-5’ 3’-CCCCTATGCACGGGG-5’ 3’-CCCCTATGCACGGGGAGGGG-5’ 3’-CCCCTATGCACGGGGAGGGGTGGATGG-5’

l-5 l-10 1-15 l-20 l-27

Figure 2. The nucleotide sequence requirement of Xterminal bases for priming in vitro. All synthetic oligonucleotides for complex-formation assay were synthesized on a Cyclone DNA synthesizer (Milligen/Riosearch) and were purified by electrophoresis through a 20% (w/v) acrylamide/7 M-urea gel as described previously (Yoo & Ito, 1991). Cell extracts (PRDI DNA polymerase-terminal protein complex) were prepared as described previously (Yoo & Ito, 1989). The reaction mixture (50 ~1) contained 50 mw-Hepes (pH 7.6), 3 mM-dithiothreitol, 5 mM-MgCl,, 2 mM-ATP. 95 pM-[a-32P]dGTP (3000 Ci/mmol), 1 nmol of template DNA, and 20 ~1 of cell extract. After incubation at 25°C for 30 min. 5 units R&l DNase (Promega) were added to the reaction mixture. The mixture was incubated at 37°C for 30 min. The reaction was terminated by the addition of 50 ~1 of stop solution (918 M-sodium pyrophosphate, (PO5 M-EDTA, 20% (v/v) trichloroacetic acid) and incubated on ice for 5 min. After centrifugation, the supernatant was discarded and the pellet was dissolved in 30~1 of sample buffer (01 M-Tris.HCl (pH 6%), 2% (w/v) SDS, 20% (v/v) glycerol, 100/l (v/v) /l-mercaptoethanol, 9005% (v/v) b romophenol blue). The reaction products were examined on 129, SDS/polyacrylamide gel electrophoresis followed by autoradiography. Location numbers indicate distance from the left 3’ end of the PRDl genome. Oligonucleotides (a) (poly(C27)) and (b) (poly(G27)) were used as negative controls. used as templates

Initiation.

with synthetic oligonucleotides containing various additions at the 3’ end

In order to obtain information on the priming site, a series of templates containing additional bases at t,he 3’ end were made and used for the priming reaction. As shown in Table la templates 14, 15. 16 and 17, which contain additional C residue(s), showed as much priming activity as wildtype (template 18), suggesting that a few additional C residue(s) at the 3’ end do not change the priming activity of single-stranded templates. Since the nucleotide linked to the terminal protein during

initiation of PRDl DNA replication is dGMP, there are two possible explanations for the activity of the templates with additional C residues at the 3’ end: first, that the terminal C residue is used for priming no matter how many C residues exist at the 3’ end, and second, that only the specific (I residue that is the first base of the PRDl genome is used for priming and additional C residues do not affect the priming process. To test these two possibilities, three more templates were made and used in a priming assay. The results (Table l), show that templates with additional residues CG (template 19), GC (template 20), and CC (template 21; positive

130

S.-K.

Yoo and J. lto

Table Priming

activity

Oligonucleotide template

1 2 3 4 5 6 7 8 9

10 11 12 13

14 15 16 17 18

1

of synthetic

oligonucleotides

Priming activit,yt

Sequence

3’-CCCCT 3’-CCCCT

ATGCA CGGGG ATGCA CCCCC 3’-CGGGA ATGCA CGGGG 3’-CGGa ATGCA CCCCG 3’-CCCCT ATGCA CCCCC 3’-CCCCT ATGCA CGGGG 3’-CCCCT ATGCA CGGGG 3 ’ -CCCCT ----CGGGG 3’-CCCCT --G-CGGGG 3 ’ -CCCCT CGGGG -TGC3’-CCCCT AATGCAA CGGGG 3’-CCCCTAAATGCAAACGGGG 3 ’ -CCCCT w CGGGG 3’-CCCCCCCCT 3’-CCCCCCCT 3’-CCCCCCT 3’-CCCCCT 3’-CCCCT

+

AGGGG TCCCC AGGGG TCCCG AGGGG TCCCC

+ +

ATGCA ATGCA ATGCA ATGCA

CGGGG AGGGG

TGGATGG

+

CGGGG CGGGG CGGGG

TGGATGG TGGATGG TGGATGG

+ + +

TGGATGG

+ +

AGGGG AGGGG AGGGG

19

3’-CGCCCCT

ATGCA ATGCA

CGGGG AGGGG CGGGG

20 21 22 23

3’-GCCCCCT 3’-CCCCCCT 3’-ccccc 3’-GGGGG

ATGCA ATGCA ccccc GGGGG

CGGGG CGGGG ccccc GGGGG

ccccc GGGGG

ccccccc GGGGGGG

+ + -

t ,%plus ( + ) indicates that dGMP-terminal protein complex was formed. a minus I-) indicates that no complex formation was observed. A plus/minus (k ) indicates t,hat the complex formation was measurable. Mutated domains and insertions are indicated by bold capitals, deletions are marked (- ) and now specific sequences are underlined

control) supported priming as well as wild-type (template 7) in the presence of 15 bases (3’-CCCCTATGCACGGGG-5’) of the PRDl DNA replication origin, suggesting that the third base from the 3’ terminus (C residue) of templates 19, 20 and 21, which is the first 3’ base of the PRDl genome, is mainly utilized for priming. Our results showed that domains I, 2 and 3 (Fig. 1) are absolutely required for priming in vitro. Domain 4, which seems to be required for efficient in vitro replication of the double-stranded DNA template, was dispensable for priming. This finding led us to suggest that there are two distinct steps during PRDl DNA replication: first, recognition of the replication origin on double-stranded DNA and second, the priming event on single-stranded DNA. When the single-stranded DNA molecules having the origin of replication are used for complex-formation assay in vitro, the reaction probably mimics only the priming event on the 3’ single-stranded region of the double-stranded genome in viuo. Thus. it is possible that the requirements for recognition of the replication origin may be different from those for just priming. One plausible explanation for discrepancies in sequence requirements is that the whole double-stranded replication origin including domain 4 (i.e. domains 1 to 4) is very effectively

recognized as the target site for initiation by the PRDl DNA polymerase-terminal protein complex, while the specific single-stranded region (domains 1 to 3) exposed by unwinding after recognition of the replication origin by replication proteins is used for priming. This two-step theory is strongly supported by the fact that double&randed templat’es are much more efficient than single-stranded templates for the initiation reaction in vitro. Evidence based on the protein-priming mechanism, indicates that’ the 3’ strand is a true template and the 5’ strand is simply displaced during chain elongation. Thus, the order of events during protein-primed initiation of DNA replication would be: (1) binding of DNA polymerase-terminal protein to the replication origin, (2) unwinding and (3) protein-priming. If these three stages are sequentially co-ordinated to achieve precise and efficient protein-priming, it then becomes apparent that the arrangement of proteins and incoming nucleotides on single-stranded templates would not he as ef?ective as on dnublestranded templates. One interesting question about priming on singlestranded templates is “How do replication proteins exactly recognize the priming site!” The fact t.hat additional bases at the 3’ end of templates did not affect priming activity implies that’ the priming site

Communications is searched for from inside the template. The immediate candidates required as the signal for identifying t#he priming site are the two complementary domains 1 and 3. These two domains may interact with primer protein-DNA polymerase complex in order to locate the right residue, or a small secondary structure. formed via the two domains. itself serves as the signal and a certain location on t,he structure is used as the priming site. Further delicate mutational experiments would give clues to a better understanding of the roles of complementary domains during the protein-priming process. The number of protein-linked genomes that are known is growing. It, will be interesting to study the molecular details of the DNA replication mechanism in various systems and to perceive how living organisms. in nature, accomplish the ultimate goal for their survival with different strategies. We thank Drs H. Bernstein and Richard Friedman for t,heir review of the manuscript. We also thank Dan K. Braithwaite for his assistance in preparing this manuscript. This research was supported by grant GM28013 from the Nat,ional Institutes of Health and by American Cancer Society grant, (NP-704).

References Bamford, D. & Mindich. L. (1984). Characterization of the DNA-protein complex at the termini of the bacteriophage PRDl genome. J. Virol. 50, 309-315. Bamford, D., McGraw. T., Mackenzie. G. & Mindich, L. (1!$83). Identification of a protein bound to the termini of bacteriophage PRDl DNA. J. Virol. 47, 311-316. Blanco, L. & Salas. M. (1984). Characterization and purification of a phage $29-encoded DNA polymerasr required for the initiation of replication. Proc. Nat. Acad. Sci., I7.S.A. 81. 5325-5329. Bradley. 1). E. (1976). Adsorption of the R-specific bacteriophage PR4 to pili determined by a drug resistance plasmid of the W compatibility group. J. Gen. Microhiol. 95, 181.-185. Bradley. D. E. & Rutherford. E. L. (1975). Basic characterization of a lipid containing bacteriophage specific for plasmids of the P, N and W compatibility groups. Can. J. Microbial. 21, 152-163. Challberg, M. D. & Kelly, T. J., Jr (1979). Adenovirus DNA replication in, vitro. Proc. Nat. Acad. Sci., U.S.A.

76. 655-659.

(Ihallberg, M. D. & Kelly, T. J., Jr (1989). Animal virus DNA replication. Annu. Rev. B&hem. 58, 671-717.

131

Gerendasy, D. D. & Ito, J. (1987). The genome of lipidcontaining bacteriophage PRDl. which infects gramnegative bacteria, contains long, inverted terminal repeats. J. Viral. 61, 594-596. Gutierrez. J., Garmendia, C. & Salas, M. (1988). Characterization of the origins of replication of 429 DNA. Nucl. Acids Res. 16. bacteriophage 5895-5914. Hsieh, J-C., Yoo. S. & Ito, J. (1990). An essential argi nine residue for initiation of protein-primed DNA replication. Proc. Nat. Acad. hi., C’.S.A. 87. 8665-8669.

Lichy, tJ. H.. Horwitz. M. S. & Hurwitz, .J. (1981). Formation of a covalent complex between the 80,000dalton adenovirus terminal protein and 5’-dCMP in vitro.

Proc. Nat. Acad. Sci., U.S.A.

78. 2678-2682.

Olsen, R. H., Siak, J. S. & Gray. R. H. (1974). Characteristics of PRDI , a plasmid-dependent broad host range DNA bacteriophage. J. Viral. 14. 689-699. Rekosh. D. M. K.. Russell, N”. C., Bellett. A. J. D. dt Robinson, A. ,J. (1977). Identification of a protein linked to the ends of adenovirus DNA. Cell, 11. 283-295. Salas, M. (1988). Initiation of DNA replication by primer proteins: bacteriophage 429 and its relatives. Curr. 7’opics Microbial. Immunol. 136. 7 I-89. Savilahti. H. & Bamford, D. (1986). Linear DNA replication: inverted terminal repeats of five closely related Escherichia coli bacteriophages. Gene. 49. 199-205. Stillman, B. W. (1989). Initiation of eukaryotic DNA replication in vitro. Annu. Rea. Cell Biol. 5. 197-245. Stillman. B. W.. Lewis, ,J. B.. Chow, I,. T.. Mathews. M. B. & Smart. J. E. (1981). Identification of the gene and mRNA for the adenovirus t,rrminal protein precursor. Cell, 23. 497-508. Watabe. K. & Ito, J. (1983). A novel DSA polymerase induced 1)~ Bacillus subtilis phage 429. ,~ucl. Acids Res. 11, 8333-8342. Watabe, K., Leusch, M. & Ito. J. (1984). Replication of bacteriophage 429 DNA in vitro: the roles of terminal protein and DNA polymerase. Proc. Nat. Acad. rS’ci. U.S.A. 81. 5374-5378. Wides, R,. J., Challberg, M. D., Rawlins. I). It. & Kelly. T. ?J., ,Jr (1987). Adenovirus origin of DNA replication: sequence requirements for replication in vitro. Mol.

Wimmer.

Cell. Biol.

7. 864-874.

E. (1982). Genome-linked proteins of viruses. Cell, 28, 199-201. Yoo, S. & Ito. ,J. (1989). Protein-primed replication of bacteriophage PRDl genome in. vitro. L’irology, 170. 442-449. Yoo, S. & Ito. ,J. (1991). Sequence requirements for protein-primed DNA replication of bacteriophage PRDl. J. Mol. Biol. 218, 779-789.

Edited by I’. von Hippel