J. Mol. Biol. (1984) 172, 507-52] The Functional Orion of Bacteriophage fl D N A Replication Its Signals and Domains GIAN PAOLO DOTTO, KENSUKE HORIUCHI AND NORTON D. ZINDER The Rockefeller University New York, N.Y. 10021, U.S.A. (Received 13 June 1983, and in revised form 6 October 1983) The origin of DNA replication of bacteriophage fl functions as a signal, not only for initiation of viral strand synthesis, but also for its termination. Viral (plus) strand synthesis initiates and terminates at a specific site (plus origin) that is recognized and nicked by the viral gene II protein. Mutational analysis of the 5' side (upstream) of the origin of plus strand replication of phage fl led us to postulate tile existence of a set of overlapping functional domains. These included ones for strand nicking, and initiation and termination of DNA synthesis. Mutational analysis of the 3' side (downstream) of the origin has verified the existence of these domains and determined their extent. The results indicate that the fl "functional origin" can be divided into two domains: (1) a "core region", about 40 nucleotides long, that is absolutely required for plus strand synthesis and contains three distinct but partially overlappingsignais, (a)the gene II protein recognition sequence, which is necessary both for plus strand initiation and termination, (b)the termination signal, which extends for eight more nucleotides on the 5' side of the gene II protein recognition sequence, (c) the initiation signal that extends for about ten more nucleotides on the 3' side of the gene II protein recognition sequence; (2) a "secondary region", 10O nucleotides long, required exclusively for plus strand initiation. Disruption of the secondary region .does not completely abolish the functionality of the fl origin but does drastically reduce it (1% residual biological activity). W e discuss a possible explanation of the fact that this region can be interrupted (e.g. fl, M]3 cloning vectors) by large insertions of foreign DNA without significantly affecting replication. 1. Introduction The sinai[ icosahedral or filamentous single-stranded D N A bc,cteriophages such as ¢ X 1 7 4 or fI (fd, MI3) have served a s very useful model systems f o r s t u d y i n g DNA replication ( D e n h a r d t et al., 1978).A w e a l t h of information i.~ now available concerning both the e n z y m e s involved and their m e c h a n i s m s of action (Kornberg, 1980). W h a t seems still to be missing is detailed knowledge o f t h e signals p r e s e n t in t h e p h a g e genome t h a t a r e r e q u i r e d for D N A replication to specificallyinitiate a n d terminate, and how these signals are recognized a n d function. We have been u n d e r t a k i n g a detailed analysis of t h e region a r o u n d t h e origin of replication of bacteriophage fl t h a t is required for efficient viral s t r a n d synthesis t o occur 507 {X~22-2836184/040~)7-15 $03.(~qI0 © 1984 Academic Press Inc. (London) Ltd, ~-~t.~ G.P. DOTTO, K. HORIUCH! AND N. D. ZINDER (l)otto et ¢zl., 1983). The present study conclusively shows that this region contains three distinct but partially overlapping domains that are required, respectively, for initiation of DNA replication, for its termination, and for both initiation and termination. After entering the bacterial cell, the fl viral (plus} strand serves as template for the synthesis by host. enzymes of the complementary {minus} strand. I n particular, the host RNA polymerase recognizes a specific signal on the fl plus strand attd, beginning at the "minus origin", synthesizes an RNA primer of about 30 nucleotides (Geider et al., 1978). The primer is subsequently elongated by DNA polymerase IIl and eventually a double-stranded, circular, superhelicat molecule (RFI) is formed. Viral {plus} strand synthesis is then initiated by the viral gene II protein, which introduces a r~ick at a specific site {plus origin) on the plus strand of the RF] molecule {Meyer et al., 1979). Elongation of the 3' end of the nick by l)NA polymerase I l l is accompanied by displacement of the old viral strand via a rolling circle mechanism {Gilbert & Dressier, 1968). In addition to its role in initiation, gone II protein is also able, after one round of plus strand synthesis, to cleave the n~scent single-stranded tail from the replicative intermediate and seal it to form a covalent|y closed circle (Meyer & Geider, 1982). The final products, in vivo, are a closed, single-stranded molecule and a closed, double-stranded replicative form (RF) (Horiuchi et al., 1978b). In the early stages of phage infection, the single-stranded circular DNA formed serves as template for minus strand synthesis to yield more+ RF molecules. In later stages, the single strands interact with the viral gene V protein (Mazur & Model, 1973; Mazur & Zinder, 1975) and are subsequently packaged into viral particles. The phage "functional origin" of replication has been previously defined as the minimal fi m:quenee that, when harhored in a plasmid, causes it to enter the fl mode of replication provided that helper phage is present (Dotto et al., 1981a). This is manifested in three ways: (1)stimulation of plasmid RF synthesis; (2) ability of the plasmid to interfere with fl DNA replication; and (3) synthesis of plasmid single-stranded DNA that is packaged into virion-tike particles that are able to transduce resistance to antibiotics (transducing particles). Such a sequence includes the fl plus origin and extends for only 12 nucleotides on its 5' side (upstream) (Dotto et al., 1982b), but for more than 100 on its 3' side (downstream) (Dotto et al., 1981a,1982a; Cleary & Ray, 1981). (The fl minus origin, which lies about 30 nucieotides on the 5' side of the plus origin, is dispensable in this system because other sequences elsewhere in the plasmid DNA can serve for minus strand initiation (Cieary & Ray, 1981; Dotto & Zinder, l.~m3).) Deletion analysis of the 5' side of the f i functional origin has revealed that this region contains two domains essential for function: (I) the signal for initiation of plus st,rand synthesis and (2) the signal for its termination (Dot.to & Horiuchi, 1981; Dotto et al., 1982a). The 5' boundary of the signal for plus strand initiation might coincide with the 5' boundary of the gene I ! protein in vitro recognition sequence {four nue/eotides on the 5' side of the gene I I protein nicking site) while the 5' boundary of the signal for termination extends for eighb nucleotides move beyond it (Dotto el ai., 1982a). ORIGIN OF fl DNA REPLICATION 51~) In this paper, we describe several d e l e t i o n a n d insertion m u t a n t s on t h e 3' side o f the f l plus origin t h a t were c o n s t r u c t e d a n d t e s t e d for t h e i r effect on t h e ability o f the fl origin to f u n c t i o n in vitro, as a s u b s t r a t e for gene I I protein nicking a c t i v i t y , a n d in vivo, as a signal for e i t h e r initiation or t e r m i n a t i o n o f plus s t r a n d synthesis. This has led to a detailed analysis o f the 3' side o f the f l origin a n d a d e t e r m i n a t i o n o f its functions and boundaries. 2. Materials a n d M e t h o d s (a) Bacteria, phage and plasmids The bacterial strain used was Escherichia coli K38 (Lyons & Zinder, 1972). fl and M13 phages were from our laboratory stock. R283 is an fl variant with a Pet linker (G-C-T-G-C-A-G-C) inserted at position 5766 of the fl map (Dotto et al., 1982b). pBR322 (Bolivar et al., 1977), pDl2 and pDl7 (Dotto et al., 1982b) have been described, pD27 and 1)1)37 were obtained from pDl7 (a pBR derivative containing the fl HaeIIi-F fragment at its EcoRl site) by destroying either its HindIII or Aval sites by end-filling with Klenow fragment, followed by eireularization with T4 DNA ligase, pD30 is a pBR322 derivative containing the HpaII-H fragment of R283 into which a HindIII linker was inserted at its HinfI site. It was obtained as follows: DNA of pD12, which is a pBR322 derivative containing the HpaII-H fragment of R283 flanked by BamHI linkers at the BamHI site (Dotto el al., 1982b), was digested with HinfI (which cleaves the R283 fragment only once at position 5788), and the 2 fragments of interest (containing either the left side of the R283 fragment, for pD28, or the right side, for pD29, plus flanking plasmid sequences) were purified. Following end-filling with Klenow fragment, HindIII linkers (C-C-A-A-G-C-T-T-G-G) were attached and, after HindIII and BamHI cleavage, the fragments were inserted at the HindIII-BamHI sites of pBR322 (pD28 and pD29), pD30 was obtained by inserting the two R283 fragments, purified from pD28 and pD29, into the BamHI site of pD27. pD38 was obtained by inserting the Ml3 HpaII-H fragment at the BamHl site of pl)37 by use of Klenow fragment and BamHI linkers (C-C-G-G-A-T-C-C-G-G). The constructiGn of pD39, pD40 and pD48 is described in Results. (b) Plasmid and phage R F preparation The volumes given below are for a 20-ml culture. Cells containing plasmids were grown to saturation in fortified broth (2.5% (w/v) Bacto-tryptone, 0.750/0 (w/v) Bacto-yeast extract, 0.6% (w/v) NaCI, 0.1% (w/v) glucose, 50 m~-Tris,HCl, pH 7.6). For preparing phage RF, cells (K38 or derivative) were grown at 37°C to an 0.0.660 of 0.4 to 0.8 in fortified broth and infected with phage at a multiplicity of infection of 100. After 15 min at 37°C, chloramphenicol was added to 15#g/ml, and the cells were allowed to grow for another 60 to 120 min. Cells were collected by eentrifugation and resuspended in 0.75 ml of buffer A (25% (w/v) sucrose, 50 m,~-Tris- HCI, pH 8.0, 40 mM-EDTA). Lysozyme (0.2 ml of 15 mg/ml in buffer A) was added and incubated for 5 rain on ice. E D T A (0.15 ml of 0.5 M) was added and incubation was continued for 10 min. Buffer B (1 ml; 0.1% Triton X-100, 50 mM-EDTA, 50 m.~-Tris. HCI, pH 8.0) was added and incubation was continued for 10 min. It was then centrifuged at 25,000 revs/min in a type 40 rotor for 6 0 min, the supernatant was collected, and then heated to 65°C for 15 min. The floceulent.precipitate was removed by centrifuging for 15 rain at 10,000 revs/min in a Sorvall'SA-600 or 8S:34 rotor. Buffer C (1 : 3 vol.; 40% (w/v) polyethylene glycol [Carbowax 6000], 2-5 ~-NaCI) was added a n d incubated at 4°C for 2 to 4 h. The DNA was recovered by eentrifugation at 10,000 revs/min for 5 min. The supernatnnt fluid was discarded and the wet pellet was resuspended in 0-3 ml DB (DNA buffer; 10 mM-Tris-HCl, pH 8.0, 10 mM-NaCI, 0.2MEDTA): Pancreatic RNase (5/al of n 2 mg/ml solution in ,water (Worthington Bioehemieals)) was added and the mixture was incubated at 37°C for 60 min. Proteins were • 510 G. P. DOTTO, K. H O R I U C H I AND N. D. Z I N D E R removed by extraction with phenol. The aqueous phase was precipitated with ethanol and the pellet was washed once with 70°/o (v/v) ethanol and dried in vacuo. The DNA was dissolved in 100/~! DB. The yield was typically 100 gg. When neees~ry, the R F I was purified by isopycnic centrifugation in CsCl/ethidium bromide. {c) DNA manipulations All DNA manipulations have been described (Dotto et al., 1981a). Sticky end ligation of cohesive ends was carried out by use of E. coli DNA ligase (New England Biolabs) under the conditions described by the manufacturer. The linkers were purchased from Collaborative Research. (d) Deletion mutants pD38 DNA, linearized with AvaI, was treated with exonuclease Bal31 (BRL) (Gray et al., 1975): 0.4 unit of enzyme per pmol of DNA at a concentration of 0.1 pg per pl in 600 mM-NaCI, 12 mM-NaCI, 12 mM-MgSO4, 12 mM-CaCI 2, 20 mM-Tris. HCI (pH 8), 1 mMEDTA, was incubated at 12°C for 6 rain or at 30°C for 30 s. The rate of Bal31 digestion is not exactly reproducible from one experiment to another, and it was necessary to titrate the enzyme each time. The reaction was stopped by addition of EGTA, 160 mM final concentration. After eireularization with T4 DNA ligase, the DNA was again treated with Aval and then used to transform calcium-treated E. coli K38 cells. Ampicillin-resistant colonies were isolated and the plasmid DNA characterized. The nucleotide sequence of the deletions obtained from pD38 was determined as follows. The DNAs of the various mutants were cleaved with BamHI and the 3' ends were labeled by addition of Klenow fragment and [32P]dGTP (Maxam & Gilbert, 1980). The DNAs were subsequently cleaved with Hinfl "and also with TaqI in the eases where one of the two labeled pBR322 bands was expected to co-migra:'~e with the desired fl fragment. The fragments of interest were gel-purified and their sequence determined as described by Maxam & Gilbert (! 980). The procedure for determining the nueleotide sequence of A + 29 was different. A + 2 9 DNA was cleaved with AsuI and the new fragment, which corresponded to the f l - p B R fusion caused by the deletion, was gel-purified. After labeling of the 3' ends, the DNA fragment was cleaved with HpaII and the fragment of interest was gel-purified prior to nucleotide sequencing. In the case of insertion mutants pD39 and pD40, the fl H p a I I - H fragments were purified after BamHI cleavage and digested with Asul; the 3' ends were labeled and the DNA was subsequently cleaved with HaeIII. After gel purification, the sequence of the fragment of interest was determined. (e) Prod.uetion of phage and antibiotic resistance transduci~u2 particles Cells harboring the various plasmids, at a density of 2 × 108/ml, were infected with phage at a multiplicity of 20 for 20 rain; they were then washed and incubated in trypt0ne broth for 3 h. The supernatant was incubated at 56°C for 30 rain (a procedure tliat results in the killing of the cells without, inactivation of the phage) and the yield of plaque-forming units (p.f.u.) was determined. Phage stocks prepared in this way were used to infect exponentially growing cells at a multiplicity of infection of i (to minimize multiple infection). After 10 rain the infected cells were diluted into anti-fl antiserum and infective centers (i.e.) were determined by plating on K38 cells. Antiserum-treated cells were also plated as colony-forming units (e.f.u.) and incubated for 3 h at 37°C before overlaying with soft agar plus ampicillin (1 mg]plate), The number of transductants (Amp a colonies) was taken as a measure of the Amp~-transducing particles produced. (f) In vitro gene I I protein assay Gene II protein was purified asdescribed (Dotto eta/., 1982b) from E. coli K38 (pD2) cells that contain a plasm;d into which gene I I has been cloned and that produce ~large ORIGIN OF fl DNA REPLICATION 511 amounts of gene II protein (Dotto et al., 1981b). One unit of enzyme is defined by the complete conversion of 0.5/~g of fi RFI DNA'into RFII and RFIV in 20/~1 of 20 mrdTris. HCI (pH 8), 5 mM-MgCI2, 5 ms-dithiothreitol at 30oC within 30 rain. The reactions were stopped with 2/d of 20% (w/v) sucrose, 0.5% (w/v) sodium dodecyl sulfate, 200 mmEDTA, and 0 . ] ~ (w/v) bromphenol blue and analyzed by agarose gel (1%) electrophoresis in the presence of ethidium bromide (0.5 pg/ml). 3. Results (a) Isolation of deletion and insertion mutants at the f l origin R283 is an fl v a r i a n t with the H i n f I site at position 5767 (13 nucleotides u p s t r e a m of t h e plus origin) s u b s t i t u t e d by a unique Pst site (octamer). This s u b s t i t u t i o n does n o t affect the functional activity of t h e origin (Dotto et al., 1982b). Thus, the functional origin of this phage (HpaII-H f r a g m e n t , see Fig. l) has a unique HinfI site left, at position 5788, eight nucleotides d o w n s t r e a m from the plus origin. I t was of interest to see if this site could be used for s t u d y i n g t h e replicative function of t h e 3' side of t h e phage origin. F o r this purpose, we started from pD12, a pB1~322 derivative t h a t contains t h e R283 functional origin (HpaTI-H fragment), and we inserted into the H i n f I site of the R283 f r a g m e n t a H i n d I I I decamer. T h e insertion, however, completely i n a c t i v a t e d t h e . R283 origin. This was indicated by the fact t h a t t h e resulting plasmid, pD30, did n o t interfere with fl and did n o t yield t r a n s d u c i n g particles at any detectable level in the presence of helper phage (Table 1}. Therefore, pD30 could n o t be used for f u r t h e r studies. i~ ~,,--"-Hoe l~ - F , : i H~II-H ~,: HoeI~-D.--~ : I ~ e 5600 Gene .~E I • )~*.-HoeI~-G.-~:4 5800 lG [ Gene ]I Biol. ocfivity ; (~i'~. : pD30 : @$ ; pD38 - ."-+.40,56 1% I ,~+41070 1% ~ i ( ,~ Ep_~ ) i "" ~ - pD48 : @ ,~+29 : ~'. : @ : ~ ) , (0001% 100 % 1% O'OI % i Z~*ll <0001% ~) : pO39 30 % ~ : pD40 1% Fie. 1. Map of the intergenic region (IG) of bacteriophage fl (Horiuchi el al., 1978a) (upper line) and of the fl fragments present in the various plaamids (lower lines). The numbers indicate the nucleotide position on the fl map (Beck & Zink, 1981; Hill & Petersen, 1982}. (+) and (--) represent the origins of plus and minus strand syntbesis, respectively. The positions of the HaeIII-F, G, D and HpaII-H fragments are shown. The HaeIII site between F.and G is also cut by AsuI. X, Indicates the HinfI cleavage sites; ~, indicates the Aval site, present in MI3. The sites in pD30, pD39 and pD40 where extra nucleotides (10, 8, and 16, respectively) were inserted, are indicated. Blank spaces between brackets indicate the deleted sequences in A+40, 56, A+41,70 and A-t-lli The relative biological activity is as measured in Table 1. 512 G. P. DOTTO, K. HORIUCHI AND N. D. ZINDER The fl and Ml3 nucleotide sequences in the origin region as well as in the rest of the phage genome differ from each other by only a few nucleotides (Beck & Zink~ 1981; Hill & Petersen, 1982). These two phages should therefore ~'~'e interchangeable and may be used according to convenience. In particular, because of a T to G difference at position 5830, M13 rims a unique AvaI site conveniently located 40 nucleotides downstream from the plus origin (Fig. 2). For this reason, we chose to use M13 to study the 3' side of the phage functional origin. pD37 is a pBR322 derivative that has lost its AvaI site and contains the fl HaeIII-F fragment at its EcoRI site. We have shown previously that this fragment contains a signal important for virion morphogenesis ("morphogenetic signal") (Dotto et al., 1981a; Dotto & Zinder, 1983); in its presence, the yield of transducing particles from chimeras that contain a phage functional origin of replication is enhanced about 100-fold. The ability of a chimeric plasmid to yield transducing particles can be taken as a measure of the functionality of its phage origin and the presence in it of the morphogenetic signal allows a more accurate quantification of the assay (Dotto et al., 1982b). Piasmid pD38 was obtained by inserting the Ml3 functional origin (HpaII-H fragment) into the BamHI site of pD37. From pD38, by double digestion of its DNA with AvaI (cleaving only inside the M13 origin) and PvuII (cleaving inside pBR sequences, at map position 2067 (Sutcliffe, 1978)) and subsequent recircularization, a second plasmid was derived (pD48). pD48 has lost the part of the MI3 (fl) functional origin on the 3' side of the AvaI site (Figs I and 2) as well as a large fragment of the plasmid sequence next to it. In a second set of experiments, pD38 DNA was linearized with AvaI, treated with exonuclease Bal31 (Gray et al., 1975) under limiting conditions, and religated to transform calcium-treated E. coil cells. The resulting clones were screened for their ability to yield transducing particles after fl infection, pD38 was used as a standard. Several clones with absent, reduced, or nearly normal ability to yield transducing particles were chosen and the nucleotide sequences of their fl fragments were determined. Figures l and 2 show the deletion map on the 3' side of the plus origin that was obtained. In a third set of experiments, pD38 DNA was again linearized with AvaI, the ends were made blunt with Kienow fragment, and BgllI linkers (C-A-G-A-TC-T-G) were inserted before recircularization. The DNA of the resulting clones was checked for sensitivity to BgIII and also PstI. If two (or more) BgtII linkers were inserted in tandem, a new Pst site would be formed. Two plasmids, one with a new BglII site (pD39) and the other with both a new BffIII and a Pst site (pD40), were chosen for characterization by nucleotide sequencing. They were confirmed to contain at their AvaI sites a BgllI linker inserted, respectively, as a monomer (8 nucleotides, pD39) or as a dimer (16 nueleotides, pD40). (b) Plus strand synthesis ie strongly affected by the deletion~ and iusertion.s The in vivo biological activities of the deletion and insertion mutants, measured either as their ability to interfere with fl replication or to yield transducing 513 O R I G I N OF fl DNA R E P L I C A T I O N TABLE l Biological activity of the origin m u t a n t s 1'lasmid p.f.u./mlt Transducing particles per m l : ~ pD30 A+ll 1-1 x 10 ~' 1 . 4 x l 0 t~' < 1 x 10 "~ <lxl0 "7 A+29 pD48 A+40, 56 1.3x 1.3x 1.6× l-3x 4-5 x 9x 5x A+41,70 p1)39 pD40 pD38 10I= 1012 1012 1 x 108 1.1 x 101° 1.4x 10)0 1.3x 10 j° 1"5 x 1011 1 x 101° 5 x 1OI° l 0 t2 1011 101 z 101° % Relative biological activity§ . . Gene I I protein nicking[ I . . 0.01 1 1 . . Plus Plus strand strand initiation¶ termination¶ . . + + + -NT -- + NT + 1 + -- + 30 1 100 + + + -t= + + + + p.f.u., plaque-forming units; NT, not tested. 1" Phage stocks were obtained from cells harboring the various plasmids as described in Materials and Methods. :[: The number of transductants (ampicillin-resistant colonies} is taken as a measure of the ampicillin resistance transducing particles produced. I t was determined as described in Materials and Methods. § The relative biological activity of the plus origin is defined in each case as the ratio between the yield of transducing particles in the phage stock obtained with a certain deletion and that obtained with pD38, the biological activity of which is arbitrarily set at 100%. The yield of transducing particles cannot be taken as an absolute number, but. needs to be normalized to the yield of phage (transducing particles/p.f.u.) given the interfering activity of the various plasmids. II Gene II protein nicking was tested ~s described in the legend to Fig. 3. ¶ Initiation and termination of the plus strand synthesis were determined as explained in Figs 4 and 5. CCTATTGGTT G 70 G "CT T T A G,, C C60 G T T T T ,,T~T~, ,..,.---~ C~_ = ATe° = G e n e - v II protein =A fA~, G TA A C , G .~BOO"A=T"c A=T C=C C=G C=G 7oC • G T =A r _ ^M / !r ~* = ~--- ^ I-~]O ~ ("7"C T = AC ACT C A AC C T A T C T C A ABO A A A A T A A "r90 GGGCTATTCITTTTGATTT ATAA / A CAATTTATAAACAAT 5909 5900 FIo. 2. Nucleotide sequence of the fl (M13) functional origin. The numbers indicate the nucleotide positions in the fl map (Beck & Zink, 1981; Hill & Petersen, 1982}. The precise 5' boundary of the functional origin was previously determined (Dotto et al., 1982b). The 3' boundary of this region lies somewhere between positions 5867 and 5909 (Dotto et al., 1981a; Cleary & Ray, 1981). The only nucleotide by which the fl sequence differs from that of M13 in this region is indicated between brackets (position 5830). Filled arrow, gene II protein nicking site (plus origin). The deletions have been named after their relative distance from the fl plus origin. The 5' endpoints of deletions A+ 11 and A+29 are indicated. The sequence deleted in A+40, 56 and A+41, 70 is also shown. A - 4 and A - 1 2 indicate the 3' endpoints of 2 deletions on the 5' side of the plus origin already described (Dotto et al., 1982a,b). Tile open arrow on the left indicates the positio n (HinfI site) of the HindIII decamer in~rtion in pD30. The oi~.n arrow "on the right indicates the position. (At~zI site) in pD39 and pD40 where 8 or 16 nucleotides (Bf/lII linkers), respectively, Were inserted. 7.~yphens have been omitted from the sequence for clarity and the usual convention has not been followed. 514 G. P. DOTTO, K. HORIUCHI AND N. D. ZINDER particles, were compared with those of the parental plasmid, pD38 (Table l). It is clear that two classes of deletion mutant can be distinguished according to their residual activities. With deletions A+40, 56 and A+41, 70 (which extend from the 40th to the 56th and from the 41st to the 70th nucleotide from the gene II protein nicking site, respectively) the activity drops to about 1% of the control. The same effect is seen also with pD48, where the M13 sequence extending beyond the 50th nucleotide from the gene II protein nicking site had been totally removed. A much more drastic drop in biological activity (0.01% or less residual activity) was observed with two other deletions, A+29, which extends from the 29th nueleotide from the gene II protein nicking site to inside pBR sequences (position 1567 of the pBR map), and A + l l , which extends from the l lth nucleotide from the gene II protein nicking site to about 100 nucleotides downstream (position 5890 of the fl map). An insertion of eight nucleotides at the AvaI site (pD39) has relatively little effecton the function of the origin (30~/o residual activity) whereas an insertion of 16 nucleotides in the same position (pD40) causes a drop in biological activity to about 1%. (c) The 3' boundary of the gene II protein recognition sequence The effects observed with the various deletion and insertion mutants could result from the disruption of the gene 1I protein recognition sequence. In particular, this might explain the very drastic loss of biological activity observed with some of the deletions and insertions (A+ 11, A+29, and pD30). To test such a possibility, DNAs from the various mutants were used in vitro as substrates for purified gene II protein. In addition to its nicking function, this protein possesses a closing activity so that, in vitro, in the presence of Mg 2+, it converts fl I~FI molecules into RFII (relaxed nicked circles) and RFIV (relaxed, closed circles) in approximately equimolar amounts (Meyer & Geider, 1979). When this reaction was used with all the mutants described, we found that gene II protein did not cut either deletion A + l l or insertion pD30. All the others, including deletion A+29, were nicked by gene 11 protein as efficiently as fl. The minimal amount of gene II protein necessary to convert 50% of fl RFI molecules into R F I I and RFIV was approximately the same for all the mutants {Fig. 3), other than A + 11 and pl)30. From these results we can conclude that the 3' boundary of the gene II protein recognition ~equence lies somewhere between position 5791 and 5809 (the endpoints of A+ 11 and A+29) and that. the insertion of ten nucleotides in this sequence (position 5791; pD30) can completely inactivate it. (d} The 3' boundaries of the signals for initiation and termination of plus strand synthesis The f l functional origin contains two specific signals, one for initiation and the other for termination of plus strand synthesis (Horiuchi, 1980; Dotto & Horiuchi, 1981). With chimeric plasmids containing two f l functional origins (inserted in ORIGIN O F fl D N A R E P L I C A T I O N 515 pD30 &+11 ~,+29 pD48 &40,56 A41,70 pD39 pD40 pD38 + - +- +- +- +- +- +- +- +- fl +- RF ] I " ~ RFT RF . ~ ' ~ FIo. 3. Conversion of deletion and insertion mutant R F I DNA into R F I I and R F I V by purified gene I I protein in vitro. R F I DNA (l pg) from the various m.utants, as well as from fl, was incubated with ( + ) or without ( - ) purified gene II protein (1 unit) for 30 min at 30°C in a 40/d volume reaction. The assays were performed as described in Materials and Methods. A DNA band migrating slightly faster than the plasmid R F I I DNA is visible in all samples and corresponds to some contaminant chromosomal DNA. The relative positions of RFIV and R F I I of A + 2 9 and pD48 are indicated by arrows in the respective lanes. the same orientation), synthesis of chimeric plus strand DNA is initiated, after fl infection, at either one of the two fl origins and is terminated at the other. Hence, the chimeric plasmids segregate into two replicons, each of them containing on|y one origin (Dotto & Horiuchi, 1981). The effidiency of different fl origins to function as signals for either initiation or termination of plus strand synthesis can be assessed easily with this system, provided that the two components of the chimera (A and B, Fig. 4) are distinguishable from each other by size. This system has previously allowed us to map the 5' boundaries of the signals for plus strand initiation and termination (Dotto et al., 1982a) and we have employed it here to determine their 3' boundaries. pD16 is a pBR322 derivative that contains two fl wild-type functional origins (HpaII-H fragment), one inserted at the EcoRI site (or|I) and the other at the B a m H I site (oriII) (Dotto et al., 1982a). A set of plasmids analogous to pDl6 was constructed, with oriI still wild type, but with one of the various deletion and insertion mutants described above substituting for oriII. In this system, if oriII lacks the signal for initiation but retains that for termination, then, upon f l infection, DNA synthesis should occur only from oriI t o oriII, and of the two components of the chimera, only A should be found inside t h e cell as an independent replicon. On the other hand, i f oriII lacks the:signal for termination but not for initiation, D N A synthesis should start from both oriI and oriII but should terminate only at or|I, a n d only component B a n d the starting chimera should be synthesized inside the cell. If. oriII lacks the signals for both initiation • "fl ~fl RF]] RFI fl RF117 516 G. P. D O T T O , K. H O R I U C H I // ~ N,L I ~'r" ',,\ \\ B.-. ~ ~ A N D N. D. Z I N D E R A J,) // i Fit;. 4. Map of plasmids containing two fl origins, or/I and orill (see the text). A and B., lesser and greater ares of the chimeras. The orientation of the fl fragments is indicated by thin arrows. The position of the insertions or deletions on the 3' sideof the gene II protein nicking site (thick arrows), is indicated by a blank space between brackets. > - , and -], Signals for initiation and termination of plus-strand synthesis, respectively. and termination, the only kind of plasmid DNA found inside the cell should be that of the starting chimera. Cells harboring pDl6 and infected with fl contained two new species of DNA of the expected size (Fig. 5). With cells harboring a p[asmid containing at its oriIl deletion A4. 1 i, neither component A DNA nor component B could be detected. This result was not unexpected, since deletion A 4- [ ! completely inactivates the fl origin in vivo and also blocks gene II protein recognition in vitro. With a plasmid containing deletion 54.29 at its oriII, component B DNA again could not be detected, indicating that this deletion severely impairs initiation. However, component A D.NA was synthesized in good quantity, indicating that deletion A 4. 29 retains the signal for termination. The 3' boundary of this signal lies somewhere between position 5791 and 5089 (the endpoints of A4.11 and A4.29) and might well coincide with the 3' boundary of the gene II protein in vitro recognition sequence. Similar results were obtained with plasmids containing at their oriII site either deletion A+40, 56 or A+41, 70; initiation is severely compromised (very little component B synthesized) while termination is not (component A synthesized in good amounts). The fact that some component B DNA can be detected with plasmids containing A 4.40, 56 and A + 41,70, but not A 4-29, correlates well with the fact that the latter deletion affects much more drastically the functionality of the fl origin (0.01% versus 1% residual biological activity). A situation similar to that described for A4.40,56 and A + 4 1 , 7 0 i s also f o u n d w i t h a plasmid containing at, its or/II the 16-nucleotide insertion of pD40. This insertion blocks initiation of plus strand synthesis (i.e. component B synthesis), while termination (i.e. component A synthesis) is unaffected. The eight-nucleotide insertion of pD39 has very little effect on initiation, and none at all on termination. This finding correlates well with the fact that with pD39 the biological activity drops only two-or threefold, while with p D 4 0 it drops about 100=fold. ORIGIN j~.1-11 A',29 &+40,56 &t'41,76 pD39 pD40 +- +- + - + - +-+ 517 O F fl D N A R E P L I C A T I O N I +- pD16 fl _ fl RFI C B, fl ss --A Fio. 5. Proficiency of the various insertion and deletion mutants as signals for either initiation or termination of plus-strand synthesis. Exponentially growing K38 cells (20 ml) harboring plasmids with two fl origins (see the text) were infected with fl (multiplicity of infection, 50). After 30 min at 37°C, cells were harvested and the intracellular DNA was purified as described in Materials and Methods. The DNA was analyzed on a 1% (w/v) agarose gel (in the presence of ethidium bromide; 0.5 pg/ml). A and B, Positions expected for the DNAs of components A and B of the chimeras, if replicating as autonomous molecules; C, RFI DNA of the Starting chimeras. Unlabeled arrow on the left, R F I DNA of A+29, starting chimera. The positions corresponding to fl R F I and ssDNA are also indicated, fl ssDNA co-migrates with component B RFI DNA of the various plasmids. Treatment with endonuclease S l showed that with all the chimeras, DNA migrating at this position is double-stranded (Si-resistant) and therefore must correspond to component B DNA (data not shown), ssDNA, singlestranded DNA. 4. Discussion The analysis of the fl functional origin presented in this paper has conclusively identified three distinct domains that are required for efficient plus strand replication (Fig. 6). (l) The gene II protein recognition sequence, (2) the signal for initiation of plus strand synthesis, and (3) the signal for its termination. The gene H protein recognition sequence, as determined by the in vitro nicking assay, extends from not more than four nucleotides on the 5' side of the gene II protein nicking site (Dotto et al., 1982b) t o I 1 to 29 nucieotides on its 3' side. This relatively long sequence is strongly asymmetric. It does not include the palindrome around the gene II protein nicking .site, which was originally thought to be important for gene II protein recognition, and it does not contain any other feature t h a t might explain the fact that gene II protei n reacts with fl DNA (and related plasmids) only when it is in a supercoiled form (Meyer & Geider, 1979).-In the presence of Mn 2÷, gene II protein cuts both strands of DNA.at the f l origin instead of introducing a nick (Dotto etal., 1981b). This effect was ascribed by others to two gene I I protein molecules t h a t bind .to t h e same:palindromic sequence around the plus origin on opposite strands of the,DNA molecule (Meyer & Geider, 1982)..However, deletion mutants with the 5' half of the-palindrome 51~ G. P. DOTTO, K. HORIUCHI AND N. D. ZINDER fl funclionol A origin B ( core ) V' B1 ~o 5769 Gene n prolein Termination Initiation J~o B2 ~0 ~o ~o 5909 [I---:-:--- 1 ~,,'/7"/'/~ ii I Fro. 6. The fl flmetional origin, its signals and its domains. Numbers indicate nucleotide positions on the fl map. ~=b, Gene II protein nicking site; ~ ~-, palindromic region around the gene II protein nicking site. The locations of the gene I] protein recognition sequence and of the signals for initiation and termination of plus strand synthesis am indicated. Also shown are the domains into which the fl functional origin can be divided. For more details, see the text. around the plus origin removed (e.g. A - 4 , Fig. 2) are still efficiently cleaved by gene I1 protein in the presence of Mn 2+ (Dotto, unpublished results). Therefore, even under these conditions, the region of palindromic s y m m e t r y around the plus origin is not necessary for gene II protein recognition. The sequence required for nicking by gene I I protein in vivo may well be the same as t h a t recognized in vitro. This is indicated by the fact that the 5' and 3' boundaries of the gene II protein in vitro recognition sequence seem to coincide, respectively, with the 5' b o u n d a r y of the signal for initiation of in vivo plus strand synthesis and the 3' b g u n d a r y of the signal for termination. This fact is unlikely to be coincidental and suggests t h a t the gene II protein recognition sequence is an important determinant, even if not the only one, in both initiation and termination of plus strand synthesis. The notion t h a t gene II protein is required not only for initiation but also for termination of fl plus strand synthesis has already been d e m o n s t r a t e d by previous in vitro studies (see below), and is confirmed by the present results. The signal for termination of fl plus strand synthesis extends from 12 nucleotides on the 5; side of the gene II protein nicking site (Dotto et al., 1982a) to 11 to 29 nueleotides on its 3' side. It is likely to include only two elements: (1) the gene II protein recognition sequence, discussed above, and (2) the palindrome located around the gene II protein nicking site and extending for eight nucleotides more on the 5' side of the gene II protein recognition site. The invoh,ement of gene I] protein in the termination of plus strand synthesis has been studied in vitro (Meyer & Geider, 1982). After one round of replication, gene I I protein is able to cleave the displaced single-stranded tail from the rolling circle intermediate and then seal it to form a covalently closed circle. I t is reasonable to assume t h a t g e n e II protein recognizes the same sequence to nick the R F I molecule at the onset of replication and to cleave the n a s c e n t strand after one round of synthesis. The DNA conformation, however, could be quite different in the two situations. The palindrome around the gene II protein nicking site might then be required for the gene II protein recognition sequence to assume the proper ORIGIN OF fi DNA REPLICATION 514} conformation after one round of synthesis. Alternatively, the palindrome might be required only after cleavage to bring together the 5' and 3' ends of the .singlestranded molecule for circularization. In this respect, it is important to remember that the fl gene II protein, unlike the qbX gene A protein (Eisenberg.& Kornberg, 1979), does not remain covalently attached to the 5' end of the nick it produces. i n vitro replication studies, using deletion mutants with the 5' half of the palindrome removed, should test the two possibilities. The signal for initiation includes the gene II protein recognition sequence and extends for about 100 nucleotides downstream (Dotto et al., 1982a, and the present data). Therefore, it consists of the entire fl functional origin (Dotto et al., 1981a; Cleary & Ray, 1981), excluding the 5' half of the palindrome around the gene II protein nicking site. According to the results presented here, the f l functional origin can be divided into two domains, domain A, extending up to about position 5819, and domain B, extending beyond it (Fig. 6). Domain A is the "core" of the f l origin. It contains the gene II protein recognition sequence, the termination signal and, on its 3' side, an additional sequence of about ten nucleotides that is absolutely necessary for plus strand initiation. If this ten-nucleotide sequence is deleted the functionality of the fl origin drops to almost undetectable levels (0.01% residual biological activity with A + 2 9 versus l~o with A+40, 56, A+41, 70 and pD48, see Fig. 1), even if the gene II protein recognition sequence and the signal for termination are retained. The insertion of 16 instead of eight nucleotides close to the border between domain A and B has striking effects on plus strand initiation (l~/o versus 30~o residual biological activity of the fl origin). This raises the possibility that the distance between the two domains is crucial for efficient initiation to occur. Domain B contains sequences required exclusively for plus strand initiation. Deletions in this region drastically reduce the functionality of the fl origin, but not as severely as in domain A (approximately 1o~) of residual biological activity with pD48, A÷40, 56 and A+41, 70). Domain B contains the site most frequently used for cloning in the single-stranded DNA phages (Messing et al., 1977; Zinder & Boeke, 1982). Thus, this domain can be separated in a phage into two subdomains, B1 and B2 (Fig. 6), by large insertions of foreign DNA without significantly affecting replication. Domain B1, on the 5' side, consists of a stretch of about 50 nucleotides while domain B2, on the 3' side, contains about 40 nucleotides, rich in A and T. W h y large stretches of foreign DNA can be inserted in such a crucial region of the phage genome and not affect its replicative function remains to be determined. However, recent results indicate that a mutation in one of the viral genes is necessary for replication to occur under these conditions (see below) (Dotto & Zinder, 1984). The requirement for plus strand initiation of about 1O0 nueleotides on the 3' side of the gene II protein recognition sequence might be due to a specific interaction of this region with some host, or, alternatively, some phage proteins (most likely gene II or gene V proteins, th e only fl proteins known to be involved in DNA replication (Horiuehi et al., 1978a). This second possibility is now supported by the finding that phage vectors, such as those described above, when used as helper phage instead of fl wild type, are able tc rescue chimeric 520 G. P. DOTTO, K. HORIUCHI AND N. D. ZINDER plasmids partially or totally lacking domain B b u t retaining only domain A (Dotto & Zinder, 1984). pD48, A + 4 0 , 56 and A + 4 1 , 70 are induced to replicate when infected with these phages, while A + 2 9 is not. These fl helper variants have suffered a mutation in one of their structural genes. Characterization of this surprising mutation(s) t h a t reduces the sequence required for a functional origin from a minimum of 140 nuc[eotides to one of 40, will aid in u n d e m t a n d i n g the specific interactions t h a t occur between the origin of fl and its gene products. We acknowledge the skilful technical assistance of Judith Schurko. This work was supported in part by grants from the National Science Foundation and the National Institutes of Health. REFERENCES Beck, E. & gink, B. (1981). Gene, 16, 35-58. Bolivar, F., Rodriguez, R. L., Green, P. J., Betlaeh, M. C., Heyneker, H. L., Boyer, H.W., Crosa, J. H. & Faikow, S. (1977). Gene, 2, 95-113. Cleary, J. M. & Ray, D. S. (1981). J. Virol. 40, 197-203. |)enhardt, D. T., Dressier, D. & Ray, D. S. (1978). Editors of The 8inffle-Stranded DNA Phages, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Dotto, G. P. & Horiuchi, K. (1981). J. Mol. Biol. 153, 169-176. l)otto, G. P. & Zinder, N. D. (1983). Virology, 130, 252-256. l)otto, G. P. & Zinder, N. D. (1984). Proc. Nat. Acad. Sci., U.S.A. In the press. Dotto, G. P., Enea, V. & Zinder, N. D. (1981a). Virology, 114, 463-473. Dotto, G. P., Enea, V. & Zinder, N. D. (1981b). Proc. Nat. Acad. Sci., U.S.A. 78, 5421.5424. Dotto, G. P., Horiuchi, K. & Zinder, N. D. (1982a). Proc. Nat. Acad. 8ci., U.S.A. 79, 71227126. Dotto, G. P., Horiuchi, K., Jakes, K. S. & Zinder, N. D. (1982b). J. Mol. Biol. 162, 335343. Dotto, G. P., Horiuchi, K., Jakes, K. S. & Zinder, N. D. (1983). Cold Spring Harbor Syrup. Quant. Biol. 47, 7 ] 7-722. Eisenberg, S. & Kornberg, A. (1979). J. Biol. Chem. 254, 5328-5332. Geider, K., Beck, E. & Schaller, H. (1978). Proc. Nat. Acad. Sci., U.S.A. 75, 645-649. Gilbert, W. & Dressier, D. (1968). Cold Spring Harbor Syrup. Quant. Biol. 33,473-485. Gray, H. B., Ostrander, D. A., Hodnett, J. L., Legerski, R. J. & Robberson, D. L. (1975). Nucl. Acids Res. 2, 1459:1492. Hill, D. F. & Peteraen, G. B. (1982). J. Virol. 44, 32--46. Horiuehi, K. (1980). Proc. Nat. Acad. Sci., U.S.A. 77, 5226-5229. Horiuchi, K., Vovis, G. F. & Model, P. (1978a). In The 8ingle.StraTwled DNA Phage~ (Denhardt, D., Dressier, D. S. & Ray, D., eds), pp. 113-137, Cold Spring Harbor Laboratory, ColdSpring Harbor, New York. Horiuchi, K , Ravetch, J. V. & Zinder, N. D. (1978b). Cold Spring Harbor Syrup. Quant. Biol. 43, 389-399. Kornberg, A.:(1980). D N A replication, Freeman, San Francisco. Lyons, L. B. & Zinder, N.D. (1972). Virology, 49,45-60. Maxam, A. M. & Gilbert, W. (1980). In Method8 in Enzymology (Grossman, L. & Moldave, K., "eds), vo]. 65, pp. 499-560, Academic Press, New York. Mazur, B. J . & Model, P. '(1973). J. Mol. Biol. 78, 285-309. Mazur: B. J'.'& Zinder, N. D. (1975). Virology, 68, 490-502. Messing, J., Gronenborn, B., Muller-Hill,.B. & Hofschneider, P. H. (1977). Proc. Nat. Acad. 8ci., U.S.A. 74, 3642-3646. ORIGIN O F fl D N A REPLICATION 521 Meyer, T. F. & Geider, K. (1979}. J. Biol. Chem. 254, 12642-12646. Meyer, ~". F. & Geider, K. (1982). Nature (London), 296, 828-832. Meyer, T. F., Geider, K., Kurz, C. & Seha|ler, H. (1979). Nature (London), 278, 365-367. Sutcliffe, J. G. (1978). Cold Sprinq Harbor Symp. Quant. Biol. 43, 77-90. Zinder, N. D. & Boeke, J. D. (1982). Gene, 19, 1-10. Edited by M. Gottesman