.=) 1990 Oxford University Press
Nucleic Acids Research, Vol. 18, No. 10 2881
Functional analysis of the Bacillus subtilis bacteriophage SPP1 pac site Alicia Bravo, Juan C.Alonso* and Thomas A.Trautner Max-Planck-Institut fur Molekulare Genetik, Ihnestrasse 73, D-1000 Berlin 33, FRG Received February 26, 1990; Revised and Accepted April 18, 1990
ABSTRACT Encapsidation of the DNA of the virulent Bacillus subtilis phage SPP1 follows a processive unidirectional headful-mechanism and initiates at a unique genomic location (pac). We have cloned a fragment of SPP1 DNA containing the pac site flanked by reporter genes into the chromosome of B. subtilis. Infection of such cells with SPP1 led to highly efficient packaging, initiated at the inserted pac site, of chromosomal DNA. The directionality in the packaging of this DNA was the same as observed with vegetative phage DNA. Mutagenizing the chromosomal pac insert defined an 83 base pair segment containing the pac cleavage site which is sufficient to direct phage specific DNA encapsidation. The packaging recognition signal as defined can also be utilized by the SPP1 related phages 41c, SF6 and Q15.
INTRODUCTION One of the most remarkable events in the life cycle of a virus within an infected cell is the packaging of its nucleic acid into progeny virus particles. Among the many facets of this process, the selective recognition of virus DNA within a universe of host DNA has attracted interest as an intriguing case of highly specific protein-DNA interaction. Studying bacteriophage packaging, it was realized that in many instances specificity of packaging is provided by the recognition by the phage coded packaging machinery of a 'signal sequence' within replicating DNA (1). Such signal sequences occur only once per genome. In the case of phage X, packaging is initiated by the binding of a phage-encoded nuclease (terminase) to cosB. At least in vitro, the binding of the terminase at cosB generates a protein-DNA complex and introduces staggered nicks at the nearby cosN site (a region of dyad symmetry) (2, 3). Cleavage of such sequences within the concatemer generates unit-length molecules carrying identical genetic information which are encapsidated into the phage proheads. In this case and those of other 'unit-length' DNA phages (T3, T7), the size of the DNA to be packaged is determined essentially by the distance between neighboring packaging signals within concatemeric DNA (4). In other phages (TI, P1, P22, SPP1), whose DNA also goes through a concatemeric replication intermediate, discrete *
To whom correspondence should be addressed
EMBL accession no. X52481
packaging signals were also identified (5-1 1). Here, however, the packaging signal only serves to initiate packaging by the formation of a double strand cleavage at the pac site. Starting from this initial cleavage site, packaging proceeds processively and unidirectionally, leading to the encapsidation through further cleavages at non defined sites, of amounts of DNA which are slightly bigger than one genome. The length of the packaged DNA in these phages is not determined by the unit genome length spacing of packaging signals in the DNA but by parameters of the packaging machinery, possibly including the packaging capacity of the phage head. This processive headful mechanism leads to the generation of terminally redundant and partially circularly permuted DNA molecules (1,12). A highly degenerate version of this mode of packaging is the case of bacteriophage T4, where also a processive mode of DNA packaging occurs, which produces a population of terminally redundant and circularly permuted DNA molecules, but where a specific initiation site for encapsidation could not be identified (13). In this communication we are concerned with the packaging signal of the virulent B. subtilis phage SPPl. We have inserted fragments of SPP1 DNA containing the pac cleavage site sequence (9) into the chromosome of B. subtilis. DNA flanking the pac insert is packaged into phage particles following infection with SPP1, showing that the phage encoded packaging machine is interacting in trans with the packaging signal placed in the chromosome. By following the packaging of chromosomal DNA flanking the pac region, after infection of cells carrying the insert, we were able to establish the minimal sequence required for directional packaging of SPP1 DNA.
MATERIALS AND METHODS Bacterial strains and phages B. subtilis strain YB886 (14) and its derivatives bearing the blepac-cat cassette were used. A DNA segment from the SPPI EcoRl fragment I containing pac [coordinates 1 to 891 of Deichelbohrer et al. (9)] was flanked by the ble [coordinates 1548 to 2081 of McKenzie et al. (15)] and cat [coordinates 610 to 2013 of Horinouchi and Weisblum (16)] genes from pUBI10 and pC194 respectiveley (ble-pac-cat cassette). A chromosomal DNA region with a low density of EcoRI restriction sites (add region) (J. Kooistra, pers. comm.) was selected for integration of the
2882 Nucleic Acids Research, Vol. 18, No. 10
pac cleavage site Ba
X
H
4-
I1
1I
I
TX B A I0 Im
aI
ll
l--CZMii2/o
I
qzzzzzZZ777,1
l
I
I
I
I
I
I
-
--
I I I--I
I ..............
aI
L--777,-l I I. bmmj.
ble
=-,7
-
Clone 40 35 36 37 38 39 34
pac activity +
+
+
cat
Fig. 1. Physical map of the ble-pac-cat cassette and its deletion derivatives. Thin lines represent phage DNA. SPP1 deleted fragments were generated in vitro by restriction enzyme digestion of clone 40. The size of the SPP1 DNA in the cassettes is 891 bp (clone 40), 455 bp (35), 197 bp (36), 172 bp (37), 484 bp (38), 865 bp (39) and 840 (clone 34). Stippled bars represent pUBl 10 DNA containing the ble gene. Striped bars denote pC194 DNA carrying the cat gene. The arrows indicate direction of transcription. Abbreviations: A, AhaII; B, BsmI; Ba, BamHI; H, HindII; T, 7haI; X, XmaIII.
ble-pac-cat cassette. For this, the cassette was joined to an 1.4 kb segment of the add chromosomal DNA region and introduced into YB886 competent cells by transformation. Chloramphenicol (Cm) resistant transformants were selected and the cassette subjected to DNA amplification as described by Young (17) using phleomycine (Pm) as selective marker (strain BG40). In vitro deletions into the SPP1 region on a plasmid-based ble-pac-cat cassette (clone 40) were performed. As revealed in Figure 1, clones 35, 36 and 37 lack the 436 bp BamHI-Hindll, 694 bp BamHI-TaI or 720 bp BamHI-XmnaH fragments respectively. In clone 35 both sites were destroyed by construction, whereas in clones 36 and 37 the BamHI and XmalII sites, respectively, are present. For the construction of clones 39 and 34, the 26 bp ThaI-XnalI and 49 bp BsmI-AhaII DNA segments, respectively, were deleted. In both constructs the restriction sites involved were destroyed. Clone 38 has an internal XMafI deletion (407 bp). The deletions were verified for the presence or absence of landmarking restriction sites. The ble-pac-cat cassettes were then integrated and amplified as described above. The clone number gives the name of the strain (i.e. clone 35 gives raise to strain BG35). Bacteriophages SPP1 (18), 41c, Q 15 and SF6 (19) were used. Transformation and media The method of Rottlander and Trautner (20) was used for transforming B. subtilis competent cells. TY broth (21) was used throughout as liquid and solid medium. It was supplemented with 5 yg/ml of Cm or 0.2 lsg/ml of Pm.
DNA techniques DNA packaged into phage particles was purified as follows: cultures of B. subtilis cells carrying the ble-pac-cat cassette were grown in TY supplemented with Pm, at 370C, to an OD560=0.8 and infected at a multiplicity of 5 phage/cell. The lysates were then treated with DNAse (1 14g/ml) and RNAse (1 ,tg/ml). Phage
particles were precipitated with PEG 20%, 2M NaCl (v/v). The DNA packaged was purified by extraction with phenol and dialyzed against 10 mM Tris-HCl, 1 mM EDTA pH 8.0. DNA preparations digested with the appropiate restriction enzymes were electrophoresed overnight in 0.8% agarose. Southern transfer to nylon membranes (GeneScreen, NEN Research products) and hybridization to 32P-labeled DNA probes were carried out as described by Maniatis et al. (22). The ble probe is a 0.55 Kb EcoRI-XbaI fragment containing the ThaIHaeIII sequence from pUB 110 [coordinates 1548 to 2081 of Mckenzie et al. (15)]. The cat probe is a 0.73 Kb HindIJI-NcoI fragment carrying the HpaII-NcoI sequence from pC194 [coordinates 969 to 1688 of Horinouchi and Weisblum (16)]. The probes were end-labeled with polynucleotide kinase according to the procedure of Maniatis et al. (22). DNA was visualized by autoradiography with Kodak XAR film. The relative amount of DNA present in any particular band identified either by staining with EtBr or Southern hybridization was determined by density scanning of films using a laser densitometer (LKB Ultro Scan XL). To determine the degree of amplification of the ble-paccat cassette on the chromosome, DNA was digested with an enzyme that cleaves once within the amplified unit and the ratio between the amount of the amplified DNA segment with respect to the non-amplified one, identified by Southern blot using the cat probe, was determined as described above. Restriction endonucleases and DNA modifying enzymes were purchased from Boehringer, Mannheim (FRG) or from BRL, Gaithersburg, Md. (U.S.A.) and were used as specified by the manufacturers.
RESULTS The SPP1 pac region is recognized as a signal for packaging when present in the B. subtilis chromosome To analyze the requirements of a pac signal to serve as the initiator of packaging, it is desirable to place this signal in trans
Nucleic Acids Research, Vol. 18, No. 10 2883 A 2
1
35
34
no
3
4
6
5
36 7
8
2
35
34
no
3
4
5
1
2
39 3
5
4
no
40 6
7I B
9 10
1
_.
C
B 38
37 9 '0 11 '2
6
36 7
8
0
37 9
10 11
12
38 2
39 3
4
40 5
6
no
7
8
0
10
do__a a-.
Fig. 2. Analysis of DNA packaged into SPP1 particles by agarose gel electrophoresis (A and B) and Southern blot (C and D) using the cat probe. DNA from SPP1 particles grown on bacterial strains, carrying in the chromosome the SPP1 cassettes indicated or no insert, was either digested with EcoRI (odd numbered lanes) or non-digested (even numbered lanes). The arrows indicate the position of reference fragments of 3.8 Kb (lane 1) and 2.1 Kb (A: lane 12 and B: lane 10) containing the cat gene. on a replicon whose packaging following phage infection can easily be monitored. Stemnberg and Coulby (6) achieved this situation in their studies on P1 packaging by lysogenizing X phage, into which the P1 packaging signal had been inserted. In our case, an 891 bp fragment from the SPP1 EcoRI fragment 1 containing the pac cleavage site was flanked by specific markers, the ble and cat genes (Fig. 1). The resulting cassette (ble-pac-cat) was linked to defined B. subtilis DNA and integrated, by chromosomal transformation, into the B. subtilis chromosome in the vicinity of the addA gene. In the presence of 0.2 ,Ag/ml of Pm in the growth media this cassette was amplified about 6 fold (strain BG40, see Materials and Methods). No chromosomal EcoRI DNA fragments, containing the amplified cassette, smaller than 35 kb should be expected. Strain BG40 was infected with SPPl, and after lysis the DNA packaged into SPP1 particles was purified. Part of this DNA was digested with EcoRI, which neither cuts the cassette ble-pac-cat nor the chromosomal DNA surrounding it. Such DNA was
analyzed by agarose gel electrophoresis and Southern hybridization using as probes end-labeled restriction fragments from both the ble (ble probe) and cat (cat probe) genes. Within the standard EcoRI restriction pattern of SPPI DNA an additional band, which comigrates with undigested phage DNA, was detected (Fig. 2B, lane 7). This material represents about 1 % of the total DNA packaged, as estimated by densitometry of the EtBr stained DNA. Only this EcoRI resistant DNA specifically hybridizes with the cat probe (Fig. 2D, lane 7). Identical results were obtained when the ble probe was used (data not shown). Therefore, such DNA packaged into the SPPI particles must derive from the amplified cassette integrated into the chromosome. When the encapsidated DNA from phages grown on strain BG40, was digested with AflJ (this enzyme cuts bacterial DNA but not phage DNA) a band which hybridizes with the cat and ble probes was detected (see below). These results indicate that chromosomal DNA adjacent to the pac region is encapsidated and that the presence of the 891 bp SPPI fragment in the B.
2884 Nucleic Acids Research, Vol. 18, No. 10
-,6
.11j
.i;.4
1:4
i,
I (.
/41 4 1 p15^ pl5... r--S- F 6 SPP1 1 2 3. L4 r-, b 7 F":""
12
j/ -IF.-
...
3 '21 k .-
& 0
o
46
0
0
0
Is
0
3.1 kb - .k4 b 2 .4 5 k b
'.5"..k.29kb 3635L43e
' .5 K( E ' k, ,
Al 6
r
e
o
ia
i. . . . b. . . t -A i3
Fig. 3. Directionality of chromosomal DNA packaging. Upper panel: DNA from SPP1 particles grown on strains BG34, BG35 or BG36 was digested with AflIl, separated on 0.8% agarose and blot hybridized with either the ble probe or the cat probe, as shown. The lengths of two restriction fragments used as references for hybridization are indicated at the left of the gel. Lower panel: scheme of the structure generated in the chromosome of the strains used in this work. Stippled bar denotes the ble gene, striped bar denotes the cat gene, the open box indicates phage DNA containing the pac site (closed circle), thin line represents the chromosomal DNA fragment (1.4 Kb) used to place the ble-pac-cat cassette in the chromosome. The AflJ (A) restriction sites are indicated. The size of the Aflll DNA fragment containing the pac site is indicated for the different strains. The cat probe does not include DNA from the smaller AlfJl DNA fragment. The arrows show the defined terminus of chromosomal DNA molecules which are expected to be present into SPP1 particles if chromosomal DNA packaging proceeds from the SPP1 pac site either to the left or to the right. The length of the putative Aflll-pac end fragments in strains BG34, BG35 and BG36 is indicated. For simplicity, in the drawing a threefold amplification of the cassette is assumed.
subtilis chromosome is sufficient for specific packaging of chromosomal DNA. Identical results were obtained when the ble-pac-cat cassette was not subjected to amplification. However, in this case the signal of chromosomal DNA packaged into SPP1 particles detected by Southern-blot was very weak (data not shown). The hybridizing band is, as expected, smaller (ranging from 15 Kb to 20 Kb depending on the strain under assay) due to the lack of amplification. This suggests that the pac site, as in the phage concatemer, is used only once. Hence, to avoid the detection problem we used in all cases strains carrying amplification of the cassette. A functional pac site is located within an 83 bp region To narrow down the size of a functional pac recognition site, deletions in the 891 bp region of SPP1 DNA were generated in vitro (Fig. 1). This involved deletions both at the right (clone 34) or at the left of the pac cleavage site sequence (clones 35, 36, 37, 38, and 39). The resulting cassettes were placed in the
Fig. 4. Recognition of the SPP1 pac site by SPP1 related phages. DNA from phage particles, obtained after infection of strain BG40 with the indicated phages, was either digested with EcoRI (lanes 1, 3, 5), or nondigested (lanes 2, 4, 6, 7), separated on 0.8% agarose and blot hybridized with the cat probe. DNA of these phages is sensitive to EcoRI.
B. subtilis chromosome at the same position as in strain BG40. The newly generated strains were termed BG34, BG35, BG36, BG37, BG38 and BG39 respectively (see Materials and Methods). DNA packaged into SPP1 particles grown on these strains was purified, digested with EcoRI and analyzed by agarose gel electrophoresis and Southern hybridization as described before. As in the DNA preparation from SPP1 particles grown on the strain BG40 (Fig. 2), an EcoRI resistant fragment which comigrates with undigested phage DNA and hybridizes with both probes was detected in the DNA preparations from strains BG34, BG35, BG36 and BG39 but not from particles obtained after infection of strains BG37 and BG38 (Fig. 2). Relating the location of the deletion endpoints within the various cassettes (Fig. 1) to the capacity of the DNA to initiate encapsidation, we realize that the 83 bp ThaI-BsmI DNA segment is shared by the packable clones 40, 35, 36, and 34. Since clones 37 and 38 are not packable, these findings would place the packaging recognition signal within this sequence. The fact that the cassette of clone 39 is recognized by the SPPI machinery might indicate that SPP1 sequences to the left of the ThaI site can substitute for the deletion of material between XmaIIl and ThaI. The amounts of non digestable DNA are somewhat different in the various positive clones shown in Fig. 2. We ruled out that such differences are due to a transfer defect, since they are highly reproducible, but we cannot say whether such differences reflect different efficiencies of packaging or whether they are due to variations of amplification or packaging initiation among the clones analyzed.
Nucleic Acids Research, Vol. 18, No. 10 2885 GENOME
~~~~~~~~~~:....
...
:
pac
-D
DIRECTION OF MATURATION
HEADFUL 104%
I\
E
II
1.111
llle
12i I
1
4
5
13
It 2
It
15 7 11
6
3
ill 16 912
-_~~~~~~~~~~~~~~~r
: 695
T5
G6
5'CGGACGTGCG
GCAGAC
1410 8
sI l+i
CATGTTCTCCGCCGCL TAGTCGAA
T5 ----I ---n777 CTATAAC CCGIT CCGC;TT GT I__, _ J
Fig. 5. Proposed processive headful mechanism for SPPl DNA maturation. The sequential and unidirectional DNA packaging from thepac end starts by recognition and cleavage of the pac site (closed circle) on a concatemer. The EcoRI restriction map of the first encapsidated DNA molecule is shown. Numbers indicate the EcoRI fragment positions on agarose gel. The right end (pac end) is generated by specific cleavage at the pac site, while the left end is generated by a nonspecific cleavage (headful cut). The nucleotide sequence of the 83 bp 7haI-BsmI region including the pac site is shown. The thickness and length of the vertical arrows indicate the frequency of cutting at the pac cleavage site.
Chromosomal DNA is packaged unidirectionally from the SPP1 pac site To determine whether the 83 bp SPPI region defines directionality of packaging and is used only once, as in the phage concatemer, DNA from SPP1 particles grown on bacteria carrying the pac site in the chromosome, was digested with Aflll and analyzed by Southern hybridization using both the cat and ble probes (Fig. 3). No restriction sites for the Afm enzyme are present in the SPP1 DNA. If packaging of bacterial DNA starts at any SPP1 pac site within the amplified region and proceeds, from such a point, unidirectionally in the same direction reported for the phage DNA (from right to left in Fig. 3), we expect, in addition to an Aflll fragment which would hybridize with both probes, a minor fragment present in non-stoichiometric amounts which would hybridize only with the ble probe. If the bacterial DNA is packaged in the opposite direction, then a minor band which would hybridize only with the cat probe should be detected. If cutting would occur at all pac sites within the amplified region, we expect only a DNA fragment which would hybridize either with the ble or cat probe. We analyzed DNA from SPP1 particles grown on the strains BG34, BG35 or BG36. AMlI restriction fragments of 3.1 Kb, 2.7 Kb or 2.45 Kb respectively, which
hybridize with both probes, were detected (see Fig. 3). In addition to these major bands, minor fragments of 1.9 Kb, 1.5 Kb or 1.2 Kb, respectively, were observed only when the ble probe was used. These results indicate that chromosomal DNA is packaged unidirectionally from the SPP1 pac site and in the same direction reported for the phage DNA (toward the EcoRI fragment 12, see Fig. 5). In addition, we conclude from those data that the pac site present in the chromosomal amplified region is utilized only once.
The SPP1 pac site is recognized by SPP1 related phages By various criteria the B. subtilis bacteriophages 41c, Ql5 and SF6 are highly related to SPP1 (19, 23, 24). To determine whether these phages recognize the SPP1 pac site as a signal for packaging, strain BG40 was infected with the phages 41c, QlS or SF6 and, after lysis, the DNA packaged into phage particles was purified and part of it was digested with EcoRI. Analysis of such DNAs by Southern hybridization using the endlabeled cat probe shows that also these phages are able to specifically encapsidate chromosomal DNA (Fig. 4). Quantification of the amount of this chromosomal DNA, by densitometric scanning of the autorradiogram, revealed that phages 41c and SF6 recognize the SPP1 pac site with an efficiency similar to SPP1. In contrast, phage Qi5 recognizes the SPP1 pac site with a 3-fold lower efficiency.
DISCUSSION To facilitate an analysis of the DNA packaging mechanism by bacteriophage SPP1 we have engineered a fragment of SPP1 DNA, containing the previously identified pac cleavage site (9) into the chromosome of B. subtilis. Following infection of such cells by SPP1, this inserted phage DNA serves as an initiation site for encapsidation of chromosomal DNA, indicating that the SPPI insert carries the recognition sequence for the SPP1 packaging machinery. Systematically reducing the size of the inserted SPP1 DNA, we could allocate the capacity of the DNA to become recognized to an 83 bp subfragment of the original insert. The nucleotide sequence of the 83 bp region shows some peculiar features (Fig. 5). Located towards the middle of the sequence (coordinates 722 to 726) are the major and the minor pac cleavage sites, which had been described before (9). The cleavage sites fall into a pentanucleotide sequence GCCGC which is repeated within the 83 bp fragment at coordinates 758 to 762 and 767 to 771. This sequence is the 5' end of a decanucleotide sequence GCCGCAATAG which is repeated beginning at coordinate 758. Furthermore, sequences TTTTT are located symmetrically around the pac cleavage site contained in the decanucleotide. It is difficult to assess the significance of the nucleotide sequence motifs and their arrangement in recognition and cleavage at the pac site. Most likely the minimal length for pac recognition and cleavage could be less than 83 bp. The use of fragments smaller than 83 bp in our packaging assay, produced by systematic exonuclease digestion from the ThaI and BsmI ends, could lead to an accurate determination of the minimal size for pac utilization. Comparing the SPP1 pac site with that of P1, as defined by Stemnberg and Coulby (6, 7), we realize that a highly structured DNA sequence is also involved here. The minimal length of P1 DNA required for packaging and its directionality amounts to about 160 bp. Also here the pac cleavage site is rather
2886 Nucleic Acids Research, Vol. 18, No. 10
symmetrically imbedded in a series of seven direct repeat sequences of 6 bp. It is of interest that both in the SPP1 and P1 cleavage site the same nucleotide sequence GCCGC is involved. The B. subtilis phages 41c, SF6 and Q 15, are closely related to SPP1 (24). By electron-microscopic and Southern hybridization analysis, heterologous and homologous genome regions between these four phages have been defined (23). The phages 41c and SF6 recognize the SPP1 pac site as signal for packaging initiation with an efficiency similar to SPP1, while the recognition efficiency is about 3 fold lower for Q 15 when is compared with SPP1 (Fig. 4). These results suggest a higher divergence between Q 15 and SPP1 with respect to their packaging systems. Comparisons to the nucleotide level between both systems would be of particular interest for studying molecular evolution.
ACKNOWLEDGEMENTS This research was partially supported by Deutsche Forschungsgemeinschaft (Al 284/1-1). A.B. was supported by an EMBO long term fellowship.
REFERENCES 1. Casjens,S.R. (1985) In Virus Structure and Assembly (Casjens,S.R. ed.) pp. 75-147, Jones and Bartlett Publishers Inc. 2. Becker,A. and Gold,M. (1978) Proc. Nat. Acad. Sci. U.S.A. 75,4199-4203. 3. Feiss,M., Widner,W., Miller,G., Johnson,G. and Christiansen,S. (1983) Gene 24,207 -218. 4. Yamagishi,M., Fujisawa,H. and Minagawa,T. (1985) Virology
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
23. 24.
144,502-515. Ramsay,N. and Ritchie,D.A. (1983) Nature 301,264-266. Sternberg,N. and Coulby,J. (1987) J. Mol. Biol. 194,453-468. Sternberg,N. and Coulby,J. (1987) J. Mol. Biol. 194,469-479. Casjens,S.R., Huang,W.M., Hayden,M. and Parr,R. (1987) J. Mol. Biol.
194,411-422. Deichelbohrer,I., Messer,W. and Trautner,T.A. (1982) J. Virol. 42,83-90. Casjens,S.R. and Huang,W. (1982) J. Mol. Biol. 157, 287-298. Backhaus,H. (1985) J. Virol. 55, 458-465. Black,L.W. (1988) The Bacteriophages (Calendar,R. ed.) Vol. 2 pp. 321-373. Plenum Press. New York and London. Kalinski,A. and Black,L.W. (1986) J. Virol. 58,951-954. Yasbin,R.E., Field.P.I. and Anderson,B.J. (1980) Gene 12,155-159. Mckenzie,T., Hoshino,T., Tanaka,T., Sueoka,N. (1986) Plasmid 15,93-105. Horinouchi,S. and Weisblum,B. (1982) J. Bacteriol. 150,815-825. Young,M. (1984) J. Gen. Microbiol. 130,1613-1621. Riva,S., Polsinelli,M. and Falaschi,A. (1968) J. Mol. Biol. 35,347-356. Santos,M.A., de Lencastre,H. and Archer,L.J. (1984) J. Gen. Virol. 65,2067-2072. Rottlander,E. and Trautner,T.A. (1970) Mol. Gen. Genet. 108,47-60. Biswal,N., Kleinschmidt,A.K., Spatz,H.C. and Trautner,T.A. (1967) Mol. Gen. Genet. 100,39-55. Maniatis,T., Fritsch,E.F. and Sambrook,J. (1982) Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Santos,M.A., Almeida,J., de Lencastre,H., Morelli,G., Kamke,M. and Trautner,T.A. (1986) J. Virol. 60,702-707. Ratcliff,S.W., Luh,J., Ganesan,A.T., Behrens,B., Thompson, R., Montenegro,M.A., Morelli,G. and Trautner,T.A. (1979) Mol. Gen. Genet. 168,165-172.
Nucleic Acids Research, Vol. 18, No. 10 2881
Functional analysis of the Bacillus subtilis bacteriophage SPP1 pac site Alicia Bravo, Juan C.Alonso* and Thomas A.Trautner Max-Planck-Institut fur Molekulare Genetik, Ihnestrasse 73, D-1000 Berlin 33, FRG Received February 26, 1990; Revised and Accepted April 18, 1990
ABSTRACT Encapsidation of the DNA of the virulent Bacillus subtilis phage SPP1 follows a processive unidirectional headful-mechanism and initiates at a unique genomic location (pac). We have cloned a fragment of SPP1 DNA containing the pac site flanked by reporter genes into the chromosome of B. subtilis. Infection of such cells with SPP1 led to highly efficient packaging, initiated at the inserted pac site, of chromosomal DNA. The directionality in the packaging of this DNA was the same as observed with vegetative phage DNA. Mutagenizing the chromosomal pac insert defined an 83 base pair segment containing the pac cleavage site which is sufficient to direct phage specific DNA encapsidation. The packaging recognition signal as defined can also be utilized by the SPP1 related phages 41c, SF6 and Q15.
INTRODUCTION One of the most remarkable events in the life cycle of a virus within an infected cell is the packaging of its nucleic acid into progeny virus particles. Among the many facets of this process, the selective recognition of virus DNA within a universe of host DNA has attracted interest as an intriguing case of highly specific protein-DNA interaction. Studying bacteriophage packaging, it was realized that in many instances specificity of packaging is provided by the recognition by the phage coded packaging machinery of a 'signal sequence' within replicating DNA (1). Such signal sequences occur only once per genome. In the case of phage X, packaging is initiated by the binding of a phage-encoded nuclease (terminase) to cosB. At least in vitro, the binding of the terminase at cosB generates a protein-DNA complex and introduces staggered nicks at the nearby cosN site (a region of dyad symmetry) (2, 3). Cleavage of such sequences within the concatemer generates unit-length molecules carrying identical genetic information which are encapsidated into the phage proheads. In this case and those of other 'unit-length' DNA phages (T3, T7), the size of the DNA to be packaged is determined essentially by the distance between neighboring packaging signals within concatemeric DNA (4). In other phages (TI, P1, P22, SPP1), whose DNA also goes through a concatemeric replication intermediate, discrete *
To whom correspondence should be addressed
EMBL accession no. X52481
packaging signals were also identified (5-1 1). Here, however, the packaging signal only serves to initiate packaging by the formation of a double strand cleavage at the pac site. Starting from this initial cleavage site, packaging proceeds processively and unidirectionally, leading to the encapsidation through further cleavages at non defined sites, of amounts of DNA which are slightly bigger than one genome. The length of the packaged DNA in these phages is not determined by the unit genome length spacing of packaging signals in the DNA but by parameters of the packaging machinery, possibly including the packaging capacity of the phage head. This processive headful mechanism leads to the generation of terminally redundant and partially circularly permuted DNA molecules (1,12). A highly degenerate version of this mode of packaging is the case of bacteriophage T4, where also a processive mode of DNA packaging occurs, which produces a population of terminally redundant and circularly permuted DNA molecules, but where a specific initiation site for encapsidation could not be identified (13). In this communication we are concerned with the packaging signal of the virulent B. subtilis phage SPPl. We have inserted fragments of SPP1 DNA containing the pac cleavage site sequence (9) into the chromosome of B. subtilis. DNA flanking the pac insert is packaged into phage particles following infection with SPP1, showing that the phage encoded packaging machine is interacting in trans with the packaging signal placed in the chromosome. By following the packaging of chromosomal DNA flanking the pac region, after infection of cells carrying the insert, we were able to establish the minimal sequence required for directional packaging of SPP1 DNA.
MATERIALS AND METHODS Bacterial strains and phages B. subtilis strain YB886 (14) and its derivatives bearing the blepac-cat cassette were used. A DNA segment from the SPPI EcoRl fragment I containing pac [coordinates 1 to 891 of Deichelbohrer et al. (9)] was flanked by the ble [coordinates 1548 to 2081 of McKenzie et al. (15)] and cat [coordinates 610 to 2013 of Horinouchi and Weisblum (16)] genes from pUBI10 and pC194 respectiveley (ble-pac-cat cassette). A chromosomal DNA region with a low density of EcoRI restriction sites (add region) (J. Kooistra, pers. comm.) was selected for integration of the
2882 Nucleic Acids Research, Vol. 18, No. 10
pac cleavage site Ba
X
H
4-
I1
1I
I
TX B A I0 Im
aI
ll
l--CZMii2/o
I
qzzzzzZZ777,1
l
I
I
I
I
I
I
-
--
I I I--I
I ..............
aI
L--777,-l I I. bmmj.
ble
=-,7
-
Clone 40 35 36 37 38 39 34
pac activity +
+
+
cat
Fig. 1. Physical map of the ble-pac-cat cassette and its deletion derivatives. Thin lines represent phage DNA. SPP1 deleted fragments were generated in vitro by restriction enzyme digestion of clone 40. The size of the SPP1 DNA in the cassettes is 891 bp (clone 40), 455 bp (35), 197 bp (36), 172 bp (37), 484 bp (38), 865 bp (39) and 840 (clone 34). Stippled bars represent pUBl 10 DNA containing the ble gene. Striped bars denote pC194 DNA carrying the cat gene. The arrows indicate direction of transcription. Abbreviations: A, AhaII; B, BsmI; Ba, BamHI; H, HindII; T, 7haI; X, XmaIII.
ble-pac-cat cassette. For this, the cassette was joined to an 1.4 kb segment of the add chromosomal DNA region and introduced into YB886 competent cells by transformation. Chloramphenicol (Cm) resistant transformants were selected and the cassette subjected to DNA amplification as described by Young (17) using phleomycine (Pm) as selective marker (strain BG40). In vitro deletions into the SPP1 region on a plasmid-based ble-pac-cat cassette (clone 40) were performed. As revealed in Figure 1, clones 35, 36 and 37 lack the 436 bp BamHI-Hindll, 694 bp BamHI-TaI or 720 bp BamHI-XmnaH fragments respectively. In clone 35 both sites were destroyed by construction, whereas in clones 36 and 37 the BamHI and XmalII sites, respectively, are present. For the construction of clones 39 and 34, the 26 bp ThaI-XnalI and 49 bp BsmI-AhaII DNA segments, respectively, were deleted. In both constructs the restriction sites involved were destroyed. Clone 38 has an internal XMafI deletion (407 bp). The deletions were verified for the presence or absence of landmarking restriction sites. The ble-pac-cat cassettes were then integrated and amplified as described above. The clone number gives the name of the strain (i.e. clone 35 gives raise to strain BG35). Bacteriophages SPP1 (18), 41c, Q 15 and SF6 (19) were used. Transformation and media The method of Rottlander and Trautner (20) was used for transforming B. subtilis competent cells. TY broth (21) was used throughout as liquid and solid medium. It was supplemented with 5 yg/ml of Cm or 0.2 lsg/ml of Pm.
DNA techniques DNA packaged into phage particles was purified as follows: cultures of B. subtilis cells carrying the ble-pac-cat cassette were grown in TY supplemented with Pm, at 370C, to an OD560=0.8 and infected at a multiplicity of 5 phage/cell. The lysates were then treated with DNAse (1 14g/ml) and RNAse (1 ,tg/ml). Phage
particles were precipitated with PEG 20%, 2M NaCl (v/v). The DNA packaged was purified by extraction with phenol and dialyzed against 10 mM Tris-HCl, 1 mM EDTA pH 8.0. DNA preparations digested with the appropiate restriction enzymes were electrophoresed overnight in 0.8% agarose. Southern transfer to nylon membranes (GeneScreen, NEN Research products) and hybridization to 32P-labeled DNA probes were carried out as described by Maniatis et al. (22). The ble probe is a 0.55 Kb EcoRI-XbaI fragment containing the ThaIHaeIII sequence from pUB 110 [coordinates 1548 to 2081 of Mckenzie et al. (15)]. The cat probe is a 0.73 Kb HindIJI-NcoI fragment carrying the HpaII-NcoI sequence from pC194 [coordinates 969 to 1688 of Horinouchi and Weisblum (16)]. The probes were end-labeled with polynucleotide kinase according to the procedure of Maniatis et al. (22). DNA was visualized by autoradiography with Kodak XAR film. The relative amount of DNA present in any particular band identified either by staining with EtBr or Southern hybridization was determined by density scanning of films using a laser densitometer (LKB Ultro Scan XL). To determine the degree of amplification of the ble-paccat cassette on the chromosome, DNA was digested with an enzyme that cleaves once within the amplified unit and the ratio between the amount of the amplified DNA segment with respect to the non-amplified one, identified by Southern blot using the cat probe, was determined as described above. Restriction endonucleases and DNA modifying enzymes were purchased from Boehringer, Mannheim (FRG) or from BRL, Gaithersburg, Md. (U.S.A.) and were used as specified by the manufacturers.
RESULTS The SPP1 pac region is recognized as a signal for packaging when present in the B. subtilis chromosome To analyze the requirements of a pac signal to serve as the initiator of packaging, it is desirable to place this signal in trans
Nucleic Acids Research, Vol. 18, No. 10 2883 A 2
1
35
34
no
3
4
6
5
36 7
8
2
35
34
no
3
4
5
1
2
39 3
5
4
no
40 6
7I B
9 10
1
_.
C
B 38
37 9 '0 11 '2
6
36 7
8
0
37 9
10 11
12
38 2
39 3
4
40 5
6
no
7
8
0
10
do__a a-.
Fig. 2. Analysis of DNA packaged into SPP1 particles by agarose gel electrophoresis (A and B) and Southern blot (C and D) using the cat probe. DNA from SPP1 particles grown on bacterial strains, carrying in the chromosome the SPP1 cassettes indicated or no insert, was either digested with EcoRI (odd numbered lanes) or non-digested (even numbered lanes). The arrows indicate the position of reference fragments of 3.8 Kb (lane 1) and 2.1 Kb (A: lane 12 and B: lane 10) containing the cat gene. on a replicon whose packaging following phage infection can easily be monitored. Stemnberg and Coulby (6) achieved this situation in their studies on P1 packaging by lysogenizing X phage, into which the P1 packaging signal had been inserted. In our case, an 891 bp fragment from the SPP1 EcoRI fragment 1 containing the pac cleavage site was flanked by specific markers, the ble and cat genes (Fig. 1). The resulting cassette (ble-pac-cat) was linked to defined B. subtilis DNA and integrated, by chromosomal transformation, into the B. subtilis chromosome in the vicinity of the addA gene. In the presence of 0.2 ,Ag/ml of Pm in the growth media this cassette was amplified about 6 fold (strain BG40, see Materials and Methods). No chromosomal EcoRI DNA fragments, containing the amplified cassette, smaller than 35 kb should be expected. Strain BG40 was infected with SPPl, and after lysis the DNA packaged into SPP1 particles was purified. Part of this DNA was digested with EcoRI, which neither cuts the cassette ble-pac-cat nor the chromosomal DNA surrounding it. Such DNA was
analyzed by agarose gel electrophoresis and Southern hybridization using as probes end-labeled restriction fragments from both the ble (ble probe) and cat (cat probe) genes. Within the standard EcoRI restriction pattern of SPPI DNA an additional band, which comigrates with undigested phage DNA, was detected (Fig. 2B, lane 7). This material represents about 1 % of the total DNA packaged, as estimated by densitometry of the EtBr stained DNA. Only this EcoRI resistant DNA specifically hybridizes with the cat probe (Fig. 2D, lane 7). Identical results were obtained when the ble probe was used (data not shown). Therefore, such DNA packaged into the SPPI particles must derive from the amplified cassette integrated into the chromosome. When the encapsidated DNA from phages grown on strain BG40, was digested with AflJ (this enzyme cuts bacterial DNA but not phage DNA) a band which hybridizes with the cat and ble probes was detected (see below). These results indicate that chromosomal DNA adjacent to the pac region is encapsidated and that the presence of the 891 bp SPPI fragment in the B.
2884 Nucleic Acids Research, Vol. 18, No. 10
-,6
.11j
.i;.4
1:4
i,
I (.
/41 4 1 p15^ pl5... r--S- F 6 SPP1 1 2 3. L4 r-, b 7 F":""
12
j/ -IF.-
...
3 '21 k .-
& 0
o
46
0
0
0
Is
0
3.1 kb - .k4 b 2 .4 5 k b
'.5"..k.29kb 3635L43e
' .5 K( E ' k, ,
Al 6
r
e
o
ia
i. . . . b. . . t -A i3
Fig. 3. Directionality of chromosomal DNA packaging. Upper panel: DNA from SPP1 particles grown on strains BG34, BG35 or BG36 was digested with AflIl, separated on 0.8% agarose and blot hybridized with either the ble probe or the cat probe, as shown. The lengths of two restriction fragments used as references for hybridization are indicated at the left of the gel. Lower panel: scheme of the structure generated in the chromosome of the strains used in this work. Stippled bar denotes the ble gene, striped bar denotes the cat gene, the open box indicates phage DNA containing the pac site (closed circle), thin line represents the chromosomal DNA fragment (1.4 Kb) used to place the ble-pac-cat cassette in the chromosome. The AflJ (A) restriction sites are indicated. The size of the Aflll DNA fragment containing the pac site is indicated for the different strains. The cat probe does not include DNA from the smaller AlfJl DNA fragment. The arrows show the defined terminus of chromosomal DNA molecules which are expected to be present into SPP1 particles if chromosomal DNA packaging proceeds from the SPP1 pac site either to the left or to the right. The length of the putative Aflll-pac end fragments in strains BG34, BG35 and BG36 is indicated. For simplicity, in the drawing a threefold amplification of the cassette is assumed.
subtilis chromosome is sufficient for specific packaging of chromosomal DNA. Identical results were obtained when the ble-pac-cat cassette was not subjected to amplification. However, in this case the signal of chromosomal DNA packaged into SPP1 particles detected by Southern-blot was very weak (data not shown). The hybridizing band is, as expected, smaller (ranging from 15 Kb to 20 Kb depending on the strain under assay) due to the lack of amplification. This suggests that the pac site, as in the phage concatemer, is used only once. Hence, to avoid the detection problem we used in all cases strains carrying amplification of the cassette. A functional pac site is located within an 83 bp region To narrow down the size of a functional pac recognition site, deletions in the 891 bp region of SPP1 DNA were generated in vitro (Fig. 1). This involved deletions both at the right (clone 34) or at the left of the pac cleavage site sequence (clones 35, 36, 37, 38, and 39). The resulting cassettes were placed in the
Fig. 4. Recognition of the SPP1 pac site by SPP1 related phages. DNA from phage particles, obtained after infection of strain BG40 with the indicated phages, was either digested with EcoRI (lanes 1, 3, 5), or nondigested (lanes 2, 4, 6, 7), separated on 0.8% agarose and blot hybridized with the cat probe. DNA of these phages is sensitive to EcoRI.
B. subtilis chromosome at the same position as in strain BG40. The newly generated strains were termed BG34, BG35, BG36, BG37, BG38 and BG39 respectively (see Materials and Methods). DNA packaged into SPP1 particles grown on these strains was purified, digested with EcoRI and analyzed by agarose gel electrophoresis and Southern hybridization as described before. As in the DNA preparation from SPP1 particles grown on the strain BG40 (Fig. 2), an EcoRI resistant fragment which comigrates with undigested phage DNA and hybridizes with both probes was detected in the DNA preparations from strains BG34, BG35, BG36 and BG39 but not from particles obtained after infection of strains BG37 and BG38 (Fig. 2). Relating the location of the deletion endpoints within the various cassettes (Fig. 1) to the capacity of the DNA to initiate encapsidation, we realize that the 83 bp ThaI-BsmI DNA segment is shared by the packable clones 40, 35, 36, and 34. Since clones 37 and 38 are not packable, these findings would place the packaging recognition signal within this sequence. The fact that the cassette of clone 39 is recognized by the SPPI machinery might indicate that SPP1 sequences to the left of the ThaI site can substitute for the deletion of material between XmaIIl and ThaI. The amounts of non digestable DNA are somewhat different in the various positive clones shown in Fig. 2. We ruled out that such differences are due to a transfer defect, since they are highly reproducible, but we cannot say whether such differences reflect different efficiencies of packaging or whether they are due to variations of amplification or packaging initiation among the clones analyzed.
Nucleic Acids Research, Vol. 18, No. 10 2885 GENOME
~~~~~~~~~~:....
...
:
pac
-D
DIRECTION OF MATURATION
HEADFUL 104%
I\
E
II
1.111
llle
12i I
1
4
5
13
It 2
It
15 7 11
6
3
ill 16 912
-_~~~~~~~~~~~~~~~r
: 695
T5
G6
5'CGGACGTGCG
GCAGAC
1410 8
sI l+i
CATGTTCTCCGCCGCL TAGTCGAA
T5 ----I ---n777 CTATAAC CCGIT CCGC;TT GT I__, _ J
Fig. 5. Proposed processive headful mechanism for SPPl DNA maturation. The sequential and unidirectional DNA packaging from thepac end starts by recognition and cleavage of the pac site (closed circle) on a concatemer. The EcoRI restriction map of the first encapsidated DNA molecule is shown. Numbers indicate the EcoRI fragment positions on agarose gel. The right end (pac end) is generated by specific cleavage at the pac site, while the left end is generated by a nonspecific cleavage (headful cut). The nucleotide sequence of the 83 bp 7haI-BsmI region including the pac site is shown. The thickness and length of the vertical arrows indicate the frequency of cutting at the pac cleavage site.
Chromosomal DNA is packaged unidirectionally from the SPP1 pac site To determine whether the 83 bp SPPI region defines directionality of packaging and is used only once, as in the phage concatemer, DNA from SPP1 particles grown on bacteria carrying the pac site in the chromosome, was digested with Aflll and analyzed by Southern hybridization using both the cat and ble probes (Fig. 3). No restriction sites for the Afm enzyme are present in the SPP1 DNA. If packaging of bacterial DNA starts at any SPP1 pac site within the amplified region and proceeds, from such a point, unidirectionally in the same direction reported for the phage DNA (from right to left in Fig. 3), we expect, in addition to an Aflll fragment which would hybridize with both probes, a minor fragment present in non-stoichiometric amounts which would hybridize only with the ble probe. If the bacterial DNA is packaged in the opposite direction, then a minor band which would hybridize only with the cat probe should be detected. If cutting would occur at all pac sites within the amplified region, we expect only a DNA fragment which would hybridize either with the ble or cat probe. We analyzed DNA from SPP1 particles grown on the strains BG34, BG35 or BG36. AMlI restriction fragments of 3.1 Kb, 2.7 Kb or 2.45 Kb respectively, which
hybridize with both probes, were detected (see Fig. 3). In addition to these major bands, minor fragments of 1.9 Kb, 1.5 Kb or 1.2 Kb, respectively, were observed only when the ble probe was used. These results indicate that chromosomal DNA is packaged unidirectionally from the SPP1 pac site and in the same direction reported for the phage DNA (toward the EcoRI fragment 12, see Fig. 5). In addition, we conclude from those data that the pac site present in the chromosomal amplified region is utilized only once.
The SPP1 pac site is recognized by SPP1 related phages By various criteria the B. subtilis bacteriophages 41c, Ql5 and SF6 are highly related to SPP1 (19, 23, 24). To determine whether these phages recognize the SPP1 pac site as a signal for packaging, strain BG40 was infected with the phages 41c, QlS or SF6 and, after lysis, the DNA packaged into phage particles was purified and part of it was digested with EcoRI. Analysis of such DNAs by Southern hybridization using the endlabeled cat probe shows that also these phages are able to specifically encapsidate chromosomal DNA (Fig. 4). Quantification of the amount of this chromosomal DNA, by densitometric scanning of the autorradiogram, revealed that phages 41c and SF6 recognize the SPP1 pac site with an efficiency similar to SPP1. In contrast, phage Qi5 recognizes the SPP1 pac site with a 3-fold lower efficiency.
DISCUSSION To facilitate an analysis of the DNA packaging mechanism by bacteriophage SPP1 we have engineered a fragment of SPP1 DNA, containing the previously identified pac cleavage site (9) into the chromosome of B. subtilis. Following infection of such cells by SPP1, this inserted phage DNA serves as an initiation site for encapsidation of chromosomal DNA, indicating that the SPPI insert carries the recognition sequence for the SPP1 packaging machinery. Systematically reducing the size of the inserted SPP1 DNA, we could allocate the capacity of the DNA to become recognized to an 83 bp subfragment of the original insert. The nucleotide sequence of the 83 bp region shows some peculiar features (Fig. 5). Located towards the middle of the sequence (coordinates 722 to 726) are the major and the minor pac cleavage sites, which had been described before (9). The cleavage sites fall into a pentanucleotide sequence GCCGC which is repeated within the 83 bp fragment at coordinates 758 to 762 and 767 to 771. This sequence is the 5' end of a decanucleotide sequence GCCGCAATAG which is repeated beginning at coordinate 758. Furthermore, sequences TTTTT are located symmetrically around the pac cleavage site contained in the decanucleotide. It is difficult to assess the significance of the nucleotide sequence motifs and their arrangement in recognition and cleavage at the pac site. Most likely the minimal length for pac recognition and cleavage could be less than 83 bp. The use of fragments smaller than 83 bp in our packaging assay, produced by systematic exonuclease digestion from the ThaI and BsmI ends, could lead to an accurate determination of the minimal size for pac utilization. Comparing the SPP1 pac site with that of P1, as defined by Stemnberg and Coulby (6, 7), we realize that a highly structured DNA sequence is also involved here. The minimal length of P1 DNA required for packaging and its directionality amounts to about 160 bp. Also here the pac cleavage site is rather
2886 Nucleic Acids Research, Vol. 18, No. 10
symmetrically imbedded in a series of seven direct repeat sequences of 6 bp. It is of interest that both in the SPP1 and P1 cleavage site the same nucleotide sequence GCCGC is involved. The B. subtilis phages 41c, SF6 and Q 15, are closely related to SPP1 (24). By electron-microscopic and Southern hybridization analysis, heterologous and homologous genome regions between these four phages have been defined (23). The phages 41c and SF6 recognize the SPP1 pac site as signal for packaging initiation with an efficiency similar to SPP1, while the recognition efficiency is about 3 fold lower for Q 15 when is compared with SPP1 (Fig. 4). These results suggest a higher divergence between Q 15 and SPP1 with respect to their packaging systems. Comparisons to the nucleotide level between both systems would be of particular interest for studying molecular evolution.
ACKNOWLEDGEMENTS This research was partially supported by Deutsche Forschungsgemeinschaft (Al 284/1-1). A.B. was supported by an EMBO long term fellowship.
REFERENCES 1. Casjens,S.R. (1985) In Virus Structure and Assembly (Casjens,S.R. ed.) pp. 75-147, Jones and Bartlett Publishers Inc. 2. Becker,A. and Gold,M. (1978) Proc. Nat. Acad. Sci. U.S.A. 75,4199-4203. 3. Feiss,M., Widner,W., Miller,G., Johnson,G. and Christiansen,S. (1983) Gene 24,207 -218. 4. Yamagishi,M., Fujisawa,H. and Minagawa,T. (1985) Virology
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
23. 24.
144,502-515. Ramsay,N. and Ritchie,D.A. (1983) Nature 301,264-266. Sternberg,N. and Coulby,J. (1987) J. Mol. Biol. 194,453-468. Sternberg,N. and Coulby,J. (1987) J. Mol. Biol. 194,469-479. Casjens,S.R., Huang,W.M., Hayden,M. and Parr,R. (1987) J. Mol. Biol.
194,411-422. Deichelbohrer,I., Messer,W. and Trautner,T.A. (1982) J. Virol. 42,83-90. Casjens,S.R. and Huang,W. (1982) J. Mol. Biol. 157, 287-298. Backhaus,H. (1985) J. Virol. 55, 458-465. Black,L.W. (1988) The Bacteriophages (Calendar,R. ed.) Vol. 2 pp. 321-373. Plenum Press. New York and London. Kalinski,A. and Black,L.W. (1986) J. Virol. 58,951-954. Yasbin,R.E., Field.P.I. and Anderson,B.J. (1980) Gene 12,155-159. Mckenzie,T., Hoshino,T., Tanaka,T., Sueoka,N. (1986) Plasmid 15,93-105. Horinouchi,S. and Weisblum,B. (1982) J. Bacteriol. 150,815-825. Young,M. (1984) J. Gen. Microbiol. 130,1613-1621. Riva,S., Polsinelli,M. and Falaschi,A. (1968) J. Mol. Biol. 35,347-356. Santos,M.A., de Lencastre,H. and Archer,L.J. (1984) J. Gen. Virol. 65,2067-2072. Rottlander,E. and Trautner,T.A. (1970) Mol. Gen. Genet. 108,47-60. Biswal,N., Kleinschmidt,A.K., Spatz,H.C. and Trautner,T.A. (1967) Mol. Gen. Genet. 100,39-55. Maniatis,T., Fritsch,E.F. and Sambrook,J. (1982) Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Santos,M.A., Almeida,J., de Lencastre,H., Morelli,G., Kamke,M. and Trautner,T.A. (1986) J. Virol. 60,702-707. Ratcliff,S.W., Luh,J., Ganesan,A.T., Behrens,B., Thompson, R., Montenegro,M.A., Morelli,G. and Trautner,T.A. (1979) Mol. Gen. Genet. 168,165-172.
