Molecular Microbiology (1992) 6(10), 1289-1296
A system of transposon mutagenesis for bacteriophage T4 Denise L. Woodworth and Kenneth N. Kreuzer* Department of Mierobiology and Immunology and Program in Cellular and Moleeular Biology. Duke University Medical Center, Durham, North Carolina 27710. USA. Summary We have developed a system of transposon mutagenesJs for bacteriophage T4. The transposon is a plasmid derivative of Tn5 which contains the essential T4 gene 24, permitting a direct selection for transposition events into a gene 24-deteted phage. The transposition occurred at a frequency of only 10~^ per progeny phage, even though a dam~ host was used to increase transposition frequency. Phage strains with a transposon insert were distinguished from most pseudorevertants of the gene 24 deletion by plaque hybridization using a transposon-specific probe. Mapping analysis showed that the transposon inserts into a large number of sites in the T4 genome, probably with a preference for certain regions. The transposon insertions in four strains were analysed by DNA sequencing using primers that hybridize to each end of the transposon and read out into the T4 genome. In each case, a 9 bp T4 target sequence had been duplicated and the insertions had occurred exactly at the \S50 ends of the transposon, demonstrating that bona fide transposition had occurred. Finally, the transposon insert strains were screened on the TabG Escherichia coii strain, which inhibits the growth of T4 motA mutants, and a motA transposon insert strain was found.
prokaryotic counterparts (Bernstein and Bernstein, 1989); understanding a particular process in the T4 infective cycle often facilitates a parallel study in eukaryotic organisms (e.g,, see Tsurimoto and SNIiman, 1990). The nearly complete sequence of the T4 genome contains about 100 open reading frames (ORFs) of unknown function. Therefore, transposon mutagenesis would be a very useful addition to the genetic arsenal available for manipulating the T4 genome, Transposon mutagenesis is a powerful approach because transposons insert into a large number of sites and generally result in complete knockout mutations. In addition, transposon insertion events can be directly selected using a gene within the transposable element, and the mutant phenotype caused by the insertion is completely linked to the selectable marker within the transposon- The numerous advantages and applications of transposon mutagenesis have been reviewed by Kleckner etai. (1977) and Berg etal. (1989). Tn5 is a well-studied transposable element that consists of three drug-resistance genes flanked by two nearly identical copies of \S50 (reviewed in Berg, 1989), Tn5 inserts with little site specificity by a non-replicative mechanism that requires the Tn5-encoded transposase protein and inverted repeats at the ends of the element. These two repeats are referred to as the outside (O) ends of the IS50 elements; the inside (I) ends are also competent as substrates for transposition. The drug-resistance markers carried by Tn5 cannot be used to select for transposition into the genome of a lytic phage such as T4, In this communication, we describe the construction of a Tn5 derivative that can be used with phage T4, We demonstrate tbat transposition occurs from a plasmid to the T4 genome and characterize a library of transposon insertions throughout the T4 genome.
Introduction Bacteriophage T4 is an extremely useful model system for studying a large variety of biological processes, T4 grows quickly, has a relatively small genome (166 kb) which has been well characterized, and encodes many of the proteins necessary for its infective cycle. Additionally, the sequences of T4 proteins show about as much homology with their eukaryotic counterparts as with their Received 31 October, 1991; revised and accepted 28 January, 1992, "For correspondence, Tel, (919) 684 2566: Fax (919) 681 8911.
Results Design of the transposon system In developing a transposon mutagenesis system tor T4, the first important limitation was the fact that T4 DNA is extensively modified, with all cytosine residues containing glucosylated hydroxymethyl groups. It seems likely that these cytosine modifications would interfere with the transposition machinery, either preventing transposition
1290
D. L. Woodworth and K. N. Kreuzer Fig. 1. A. Ptasmid pKKE21 map. Important elements of the 7,8 kb plasmid are depicted along with the restriclion enzyme sites used for mapping. Arrows A and B indicate the oligonucleotides used for sequencing. B. Map of the plasmid dimer. Dotted lines indicate the two identical Jn5-24 transposon units that exist in a plasmid dimer.
PstI
Sail
amp
PstI
or limiting insertion events to AT-rich regions. This problem was overcome by utilizing the mutant strain T4 dC (56^"" 42^"^ (denB-rll)'^ denA ale), which carries mutations that allow the strain to produce cytosine-oontaining DNA and still survive. T4 dC can also produce modified DNA whenever an amber suppressor in the host strain suppresses the gene 56 and 42 mutations. This possibility is useful because T4 grows more efficiently with modified DNA and because screens for transposon-induced mutants can thereby be performed under growth conditions that more closely resemble those of wild-type phage. A second limitation was the lack of a suitable selective marker. T4 is a lytic phage and therefore drug-resistanoe markers characteristic of most transposons would not be effective. We used the T4 gene 24 selection system developed by H. Engman and K. N. Kreuzer (unpublished data). Phages with a deletion of gene 24, which encodes an essential head protein, are viable only when the gene 24 product (gp24) is supplied by a host plasmid. However, transposition of a gene 24-oontaining transposon into the gene 24 deletion phage would allow growth on a plasmid-free host. A T4 dC derivative containing the gene 24 deletion. T4 CD17 (H. Engman and K. N. Kreuzer, unpublished data), was available for use as a transposon recipient that can propagate with unmodified cytosine residues in its DNA.
and I ends of the IS50 elements from Tn5, the Tn5 transposase gene, and the pBR322 ampicillin resistance gene and origin of replication (Phadnis and Berg, 1987). Because of the orientation of the O and 1 ends, monomeric plasmid pBRG1306 is not a functional transposon. However, Phadnis and Berg (1987) demonstrated that dimerization of pBRGi306 results in two identical transposon units, each consisting of a monomeric unit of plasmid flanked by one O end and one I end (dotted lines in Fig. iB).The2.34kbT4 EcoRI fragment containing gene 24 was inserted into the unique EcoRI site of pBRG1306 to form plasmid pKKE21 (Fig. 1A); transposons generated from the dimer are referred to as Jn5-24 (Fig. 1B).
Transposition
The gene 24 selection system is limited by the generation ol pseudorevertants at a frequency of about 10"'' (H. Engman and K. N. Kreuzer, unpublished data). One class of pseudorevertant acquires a 'bypass' mutation in gene 23 that eliminates the need for gp24 (McNicol ef al., 1977). A second class acquires the gp^-^-providing plasmid by illegitimate recombination. We will refer to all CD17 derivatives that do not require the gp24 supply plasmid as •gp24-independent', even if they contain a copy of gene 24 in their genomes.
The Tn5-24 plasmid (pKKE21; Fig. 1 A) was transformed into a dam~ derivative of Escherichia coli ED8689. This non-suppressing host carries a mutation that inactivates the hsd restriction system that normally digests T4 oytosine-containing DNA. The dam~ mutation increases the rate of Tn5 transposition, primarily because of increased transcriptionof the transposase gene (Makris etal., 1988; McCommis and Syvanen, 1988; Yin etal., 1988; Dodson and Berg, 1989). Transformed cells wifh plasmid dimers were isolated and then infected with the T4 dC phage deleted for gene 24 (CD17), hopefully allowing transposition from the plasmid into the phage DNA. The resulting lysate was plated on suppressing E. coli cells carrying a gp24 supply plasmid (pKK032) to quantify total phage and on the same strain without the supply plasmid to select gp:?4-independent phage. A control infection was also performed using pKK032 plasmid dimers instead of the transposon plasmid. The frequencies of gp24-independent phage generated in the two infections were approximately the same (2.7 x 10"'' gp24-independent pfu per total pfu).
A gene 24-containing transposon plasmid (Fig. 1 A) was designed starting with the Tn5-derived plasmid pBRG1306. Plasmid pBRG1306 contains the 19 bp O
Although the transposon did not significantly increase the frequency of gp24-independent phage. a sizable minority of these phage had acquired a Tr\5-24 insert.
Transposon mutagenesis of bacteriophage T4
rocMoJOUrororOrorv^Q OJC\JCVICJOJC\JC\J
A system of transposon mutagenesis for bacteriophage T4 Denise L. Woodworth and Kenneth N. Kreuzer* Department of Mierobiology and Immunology and Program in Cellular and Moleeular Biology. Duke University Medical Center, Durham, North Carolina 27710. USA. Summary We have developed a system of transposon mutagenesJs for bacteriophage T4. The transposon is a plasmid derivative of Tn5 which contains the essential T4 gene 24, permitting a direct selection for transposition events into a gene 24-deteted phage. The transposition occurred at a frequency of only 10~^ per progeny phage, even though a dam~ host was used to increase transposition frequency. Phage strains with a transposon insert were distinguished from most pseudorevertants of the gene 24 deletion by plaque hybridization using a transposon-specific probe. Mapping analysis showed that the transposon inserts into a large number of sites in the T4 genome, probably with a preference for certain regions. The transposon insertions in four strains were analysed by DNA sequencing using primers that hybridize to each end of the transposon and read out into the T4 genome. In each case, a 9 bp T4 target sequence had been duplicated and the insertions had occurred exactly at the \S50 ends of the transposon, demonstrating that bona fide transposition had occurred. Finally, the transposon insert strains were screened on the TabG Escherichia coii strain, which inhibits the growth of T4 motA mutants, and a motA transposon insert strain was found.
prokaryotic counterparts (Bernstein and Bernstein, 1989); understanding a particular process in the T4 infective cycle often facilitates a parallel study in eukaryotic organisms (e.g,, see Tsurimoto and SNIiman, 1990). The nearly complete sequence of the T4 genome contains about 100 open reading frames (ORFs) of unknown function. Therefore, transposon mutagenesis would be a very useful addition to the genetic arsenal available for manipulating the T4 genome, Transposon mutagenesis is a powerful approach because transposons insert into a large number of sites and generally result in complete knockout mutations. In addition, transposon insertion events can be directly selected using a gene within the transposable element, and the mutant phenotype caused by the insertion is completely linked to the selectable marker within the transposon- The numerous advantages and applications of transposon mutagenesis have been reviewed by Kleckner etai. (1977) and Berg etal. (1989). Tn5 is a well-studied transposable element that consists of three drug-resistance genes flanked by two nearly identical copies of \S50 (reviewed in Berg, 1989), Tn5 inserts with little site specificity by a non-replicative mechanism that requires the Tn5-encoded transposase protein and inverted repeats at the ends of the element. These two repeats are referred to as the outside (O) ends of the IS50 elements; the inside (I) ends are also competent as substrates for transposition. The drug-resistance markers carried by Tn5 cannot be used to select for transposition into the genome of a lytic phage such as T4, In this communication, we describe the construction of a Tn5 derivative that can be used with phage T4, We demonstrate tbat transposition occurs from a plasmid to the T4 genome and characterize a library of transposon insertions throughout the T4 genome.
Introduction Bacteriophage T4 is an extremely useful model system for studying a large variety of biological processes, T4 grows quickly, has a relatively small genome (166 kb) which has been well characterized, and encodes many of the proteins necessary for its infective cycle. Additionally, the sequences of T4 proteins show about as much homology with their eukaryotic counterparts as with their Received 31 October, 1991; revised and accepted 28 January, 1992, "For correspondence, Tel, (919) 684 2566: Fax (919) 681 8911.
Results Design of the transposon system In developing a transposon mutagenesis system tor T4, the first important limitation was the fact that T4 DNA is extensively modified, with all cytosine residues containing glucosylated hydroxymethyl groups. It seems likely that these cytosine modifications would interfere with the transposition machinery, either preventing transposition
1290
D. L. Woodworth and K. N. Kreuzer Fig. 1. A. Ptasmid pKKE21 map. Important elements of the 7,8 kb plasmid are depicted along with the restriclion enzyme sites used for mapping. Arrows A and B indicate the oligonucleotides used for sequencing. B. Map of the plasmid dimer. Dotted lines indicate the two identical Jn5-24 transposon units that exist in a plasmid dimer.
PstI
Sail
amp
PstI
or limiting insertion events to AT-rich regions. This problem was overcome by utilizing the mutant strain T4 dC (56^"" 42^"^ (denB-rll)'^ denA ale), which carries mutations that allow the strain to produce cytosine-oontaining DNA and still survive. T4 dC can also produce modified DNA whenever an amber suppressor in the host strain suppresses the gene 56 and 42 mutations. This possibility is useful because T4 grows more efficiently with modified DNA and because screens for transposon-induced mutants can thereby be performed under growth conditions that more closely resemble those of wild-type phage. A second limitation was the lack of a suitable selective marker. T4 is a lytic phage and therefore drug-resistanoe markers characteristic of most transposons would not be effective. We used the T4 gene 24 selection system developed by H. Engman and K. N. Kreuzer (unpublished data). Phages with a deletion of gene 24, which encodes an essential head protein, are viable only when the gene 24 product (gp24) is supplied by a host plasmid. However, transposition of a gene 24-oontaining transposon into the gene 24 deletion phage would allow growth on a plasmid-free host. A T4 dC derivative containing the gene 24 deletion. T4 CD17 (H. Engman and K. N. Kreuzer, unpublished data), was available for use as a transposon recipient that can propagate with unmodified cytosine residues in its DNA.
and I ends of the IS50 elements from Tn5, the Tn5 transposase gene, and the pBR322 ampicillin resistance gene and origin of replication (Phadnis and Berg, 1987). Because of the orientation of the O and 1 ends, monomeric plasmid pBRG1306 is not a functional transposon. However, Phadnis and Berg (1987) demonstrated that dimerization of pBRGi306 results in two identical transposon units, each consisting of a monomeric unit of plasmid flanked by one O end and one I end (dotted lines in Fig. iB).The2.34kbT4 EcoRI fragment containing gene 24 was inserted into the unique EcoRI site of pBRG1306 to form plasmid pKKE21 (Fig. 1A); transposons generated from the dimer are referred to as Jn5-24 (Fig. 1B).
Transposition
The gene 24 selection system is limited by the generation ol pseudorevertants at a frequency of about 10"'' (H. Engman and K. N. Kreuzer, unpublished data). One class of pseudorevertant acquires a 'bypass' mutation in gene 23 that eliminates the need for gp24 (McNicol ef al., 1977). A second class acquires the gp^-^-providing plasmid by illegitimate recombination. We will refer to all CD17 derivatives that do not require the gp24 supply plasmid as •gp24-independent', even if they contain a copy of gene 24 in their genomes.
The Tn5-24 plasmid (pKKE21; Fig. 1 A) was transformed into a dam~ derivative of Escherichia coli ED8689. This non-suppressing host carries a mutation that inactivates the hsd restriction system that normally digests T4 oytosine-containing DNA. The dam~ mutation increases the rate of Tn5 transposition, primarily because of increased transcriptionof the transposase gene (Makris etal., 1988; McCommis and Syvanen, 1988; Yin etal., 1988; Dodson and Berg, 1989). Transformed cells wifh plasmid dimers were isolated and then infected with the T4 dC phage deleted for gene 24 (CD17), hopefully allowing transposition from the plasmid into the phage DNA. The resulting lysate was plated on suppressing E. coli cells carrying a gp24 supply plasmid (pKK032) to quantify total phage and on the same strain without the supply plasmid to select gp:?4-independent phage. A control infection was also performed using pKK032 plasmid dimers instead of the transposon plasmid. The frequencies of gp24-independent phage generated in the two infections were approximately the same (2.7 x 10"'' gp24-independent pfu per total pfu).
A gene 24-containing transposon plasmid (Fig. 1 A) was designed starting with the Tn5-derived plasmid pBRG1306. Plasmid pBRG1306 contains the 19 bp O
Although the transposon did not significantly increase the frequency of gp24-independent phage. a sizable minority of these phage had acquired a Tr\5-24 insert.
Transposon mutagenesis of bacteriophage T4
rocMoJOUrororOrorv^Q OJC\JCVICJOJC\JC\J
