Dipartimento di Biologia, Università Tor Vergata Roma, Italy1
Author for correspondence: P. Ghelardini. Tel: +39 6 72594674. Fax: +39 6 2023500. e-mail: ghelardini{at}bio.uniroma2.it
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Keywords: Bacteriophage Mu, excisionreintegration process, nucleoid folding
Abbreviations: RAGE, rapid amplification of genomic DNA ends.
a Present Address: Centro Acidi Nucleici del CNR, c/o Dipartimento di Biologia, Università Tor Vergata, via della Ricerca Scientifica, 00133 La Romanina (Roma), Italy.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
After infection, the DNA of the phage-transposon Mu integrates randomly in the host genome (for a review see Pato, 1989 ) although either hot spot or conditions are known which can bias the choice of the target site (Casadeus & Roth, 1989
; Wang & Higgins, 1994
). A peculiarity of Mu integration is the inability of the prophage to excise from the target site. In fact, during the lytic cycle, free phage DNA is never found in the cytoplasm and phage DNA extracted from the virion particles is always covalently associated with host DNA (for a review see Pato, 1989
). Therefore, after infection, the phage DNA transposes from the old integration target, i.e. from the DNA sequences flanking the virionic DNA, to a new one. Precise excision of the prophage, thereby restoring the original target sequence, has been reported with certain defective mutants or particular conditions (Bukhari, 1975
; Shapiro, 1984
) but, in all cases, the revertant strains examined lost the prophage.
In contrast with this classical description of the Mu DNA fate, we observed that mutant Mugem2ts carrying a mutation in the gem operon promoter, excises at a relatively high frequency (Ghelardini et al., 1994 , 1995
). This phenomenon, besides representing a new aspect of Mu behaviour, is peculiar since the phage DNA excised is not lost but is reintegrated at a new site in the host genome. Moreover, the reintegration site depends strictly on the original prophage localization.
To ascertain the mechanism underlying this novel recombination event, we investigated the relationship between the donor (original integration site) and the receptor (new integration site) sites implicated in the phage DNA migration during the excisionreintegration process.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
General microbiological and DNA recombinant techniques.
Standard microbiological and DNA recombinant techniques were as described by Miller (1972) and Sambrook et al. (1989)
. DNA sequences were performed with an ABI Prism 310 Sequencer (Applied Biosystems) as recommended by the manufacturer. The sequences of the synthetic oligonucleotides used in this work are as follows: 1 (target, Mu), 5'-TGGTGACAGCTTGGTTGTTATCGG-3' [used for RAGE (rapid amplification of genomic DNA ends)]; 2 (target, pUC19), 5'-GTTTTCCCAGTCACGAC-3' (RAGE); 3 (target, Mu), 5'-ATTTGCACTACAGGCTTGCAAGCC-3' (RAGE); 4 (target, pUC19), 5'-GTAAAACGACGGCCAGTGAATTCG-3' (RAGE); 5 (target, Mu), 5'-CAAATCTAATCCCATCAGATCCCG-3' (sequencing).
Excision experiments.
Construction of Mu wt and Mugem2ts lysogenic strains with the prophage inserted in a specific target gene.
Strains were constructed and proved to be monolysogens as previously described (Ghelardini et al., 1994 ).
Excision of Mugem2ts.
Bacterial strains lysogenic for Mugem2ts, with the prophage inserted in a specific target gene (lacZ, lacY or malT) were grown to early stationary phase, then centrifuged, washed three times with SM, resuspended in SM and plated on M9 medium with lactose or maltose, or isolated on MacConkey plates with the appropriate sugar to select the Lac+ (or Mal+) revertants (excised clones). Clones, grown for 23 d on M9 medium, or red clones which appeared on MacConkey plates after 35 d, were collected, purified on MacConkey plates with the appropriate sugar and stored in stabs.
Prophage localization.
The precise phage DNA insertion site was determined by RAGE (Mizobuchi & Frohman, 1993 ). Chromosomal DNA was extracted from the lysogenic strain, digested with HaeIII and ligated to pUC19 DNA digested with HincII. Five microlitres of the ligation mixture were used to perform a PCR with synthetic oligonucleotides 1 and 2 as shown above. One tenth of the PCR product volume was reamplified with nested oligonucleotides 3 and 4. The products of the reaction were separated on an agarose gel, purified and sequenced using oligonucleotide 5 as the sequencing primer.
All sequence analyses were carried out using the Wisconsin Package version 9.1, from the Genetics Computer Group (Fuchs et al., 1990 ) and FASTA.
Inversion of a large chromosomal region.
The experiments were performed and the effective inversion of the genomic region was controlled by conjugation as described by Rebollo et al. (1988) .
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Search for sequence similarities in pairs of donor/receptor sites
To test if donor and receptor sites are related in primary sequence, we analysed four receptor sites for the Mugem2ts element inserted at two donor sites within lacZ. In one case (strain R354) the phage was inserted at base 4154 with the Mu c end promoter-proximal, while in the other (strain R353), the phage was at base 3788 with the Mu S end promoter-proximal (Fig. 1a). From both strains, Lac+ revertants were obtained with frequencies around 1x10-6 and, in these clones, the prophage moved from lacZ (located at 8 min) to about 74/75 min on the E. coli genetic map (Ghelardini et al., 1995
). In particular, 21 clones, obtained from independent excision experiments with both the lacZ::Mugem2ts lysogens, resulted in a Mal- phenotype and were located in the malT gene. In these clones, the precise point of prophage insertion was determined using the method described previously (Ghelardini et al., 1994
) and the localization of eight of these Lac+ revertants, namely R362 to R366 and R378 to R380, is shown in Fig. 1b
. The data reported in Fig. 1(b)
indicate that the DNA excised from a donor site reintegrates at a cluster of points that we define as the acceptor site.
|
|
The Mugem2ts Ap1 prophage was integrated in the lacZ gene carried by the F'tslac episome of the bacterial strain R400lac to give the strain R400
lac/F'tslac::Mugem2ts Ap1 and Lac+ revertants were selected at 30 °C. As we have previously shown (Ghelardini et al., 1995
), the prophage excised from lacZ is usually reintegrated at about 74 min on the E. coli map. In this case, however, where the two sites implicated in the excisionreintegration process are in trans, on two separate DNA molecules, none of the 20 revertant clones examined genetically was located at or around 74 min. These clones were crossed with the bacterial strain R410
lac (Mu cts62) and Lac+ clones were selected. The exconjugants were then examined for the ampR marker carried by the phage, to analyse whether the prophage was still co-transferable with the episome in which it was originally inserted. Crosses, performed with 10 independent Lac+ revertants, showed that the ampR marker was co-transferred with F' and that the exconjugants acquired immunity to Mu superinfection at both 30 °C and 42 °C. Therefore, the Mugem2ts DNA, excised from the lac operon carried by F' was always reintegrated in another site of the episome itself. From these results we can conclude that the two sites involved in phage DNA migration must be located on the same DNA molecule, supporting the hypothesis that they are bound by topological constraints.
Effect of the inversion of the DNA region containing the receptor or the donor site on the Mugem2ts DNA reintegration
The strong correlation between the initial prophage location and its subsequent target, and the requirement that the two sites implicated are in cis support the hypothesis of the existence of topological constriction in Mugem2ts excisionreintegration process. At this point we may ask what would happen if the receptor site is displaced from its natural location on the E. coli map: would the target for Mugem2ts reintegration be analogously displaced or not? According to the topological-constraint hypothesis, we may expect that the phage DNA migration itself would not be modified, but, genetically, the receptor site would be changed. To answer this question, we studied the reintegration of Mugem2ts excised from the malT gene when a large inversion of the DNA region containing the phage receptor site is performed. This experiment was performed using the strategy developed by Rebollo et al. (1988) , which allows the detection of large region inversions of the E. coli chromosome.
In bacterial strain LN1883, the DNA region between tna (83 min) and purE (12 min) is surrounded by two tetA markers, both inactivated, the former with a kanR and the latter with a smR box (Fig. 3). Recombination of the two tetA genes gives rise to tetR clones in which the DNA region between them has been inverted (Rebollo et al., 1988
). Strain LN1883 was infected with Mugem2ts and lysogenic clones with the prophage integrated in the malT gene were selected. One of Mal- clones (R420) was further purified, the localization of the prophage determined as previously described (Ghelardini et al., 1994
), and tetR clones were then selected, purified and controlled for the DNA region inversion. One of the resulting clones, strain R421, was used for excision experiments.
|
Mal+ revertants were selected from R420 and from its tetR derivative R421. The prophage was localized in two independent Mal+ strains derived from R420 and in two from R421 by sequencing the junctions between the phage and the bacterial DNAs using the RAGE technique. Chromosomal DNA was digested with HaeIII, ligated to pUC19 plasmid DNA digested with HincII, amplified by PCR, and the fragments obtained were sequenced and then analysed as described in Methods.
As shown in Fig. 4, in strain R420, the prophage excised from malT is reintegrated at 91 min of the E. coli genetic map. This result is in accordance with our previous results, considering the differences between the two parental strains used and the fact that in the former case, the localization of the prophage insertion was determined genetically whereas in this case it was determined by sequence analysis. It is interesting to note that, after inversion of the genome region the reintegration site remains roughly the same distance from the donor site, i.e. after inversion, the distance between the donor and the receptor site changed from 17 to 14 min of the genetic map.
|
|
From these experiments, we can conclude that the choice of the receptor site is determined only by the location of the donor site and is independent of the nucleotide sequence of the receptor site itself. In fact, the prophage migration is not substantially affected by the inversion of the region containing the receptor site whereas the receptor site changes dramatically when the donor site is moved.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Reintegration occurs in a discrete region without any sequence recognition
One of the most intriguing peculiarities of the excision process described with the Mugem2ts mutant is that the prophage is not lost but is reintegrated at sites which depend on the original location of the prophage (Ghelardini et al., 1994 , 1995
). This prophage non-random reintegration is somehow reminiscent of the non-random pattern of Tn10-catalysed large inversions in Salmonella typhimurium, described by Krug et al. (1994)
.
Two hypotheses have been postulated to explain the relationship between the donor and receptor sites implicated in the Mu excisionreintegration process. The first suggests the existence of specific sequences at the receptor site, which are recognized and preferentially selected for the phage reintegration, for example, a sequence similarity. The second hypothesis presumes, between these two sites, the existence of some constraints that could make the two sites more reactive, for example, increasing the proximity of the two sites, which are otherwise a long way from each other along the host chromosome, or for their ability to recombine. One of these constraints may be nucleoid folding. Alternatively, we can imagine a factory model, based on that elaborated for DNA replication (Lemon & Grossman, 1998 ). With this model the two sites are brought closer together in a dynamic way at the replication-factory level by the two replication forks, which run across the donor and the receptor site, respectively.
The first hypothesis was tested by comparing the DNA sequences surrounding the Mu DNA at both the donor and receptor sites. Analysis of the receptor sites of phage DNA excised from two independent lysogens with the prophage inserted into the lacZ gene showed that they are located at 74 min on the E. coli map and that there is neither a unique site of reintegration nor a sequence similarity between the donor and receptor sites. The receptor site could be defined as a region where there is a cluster of reintegration points and may extend for a number of kilobases.
The receptor site is determined by the location of the donor site
Since Mu reintegration appears to be restricted to a discrete DNA region without any sequence recognition, we postulated that the receptor and donor sites are bound to each other by nucleoid folding or the replication factory models. Both these two models postulate that in the excisionreintegration process of Mu DNA, the receptor site should always be located on the same DNA molecule as the donor site (cis-dependence of the process). As a matter of fact, Mu DNA excised from the lacZ gene carried by the episome F' does not reintegrate in trans, at or around 74 min on the E. coli chromosome, but in cis in another region of the episome. Moreover, since F' does not contain the 74 min region, this experiment confirms that the choice of the target of the excised DNA is not determined by any nucleotide sequence, as informational content, but rather by some constraint imposed on the region itself.
In support of the sequence-independent model, we observed that the inversion of a large chromosomal region containing the receptor site does not affect the migration of the excised Mu DNA. In fact, it integrates at about the same distance from the donor site, independently of the nature of the DNA sequence. In other words, the phage DNA excised from malT integrates at 91 min in LN1883 derivative strains (i.e. at a distance of 17 min from the donor site) and at 7 min (i.e. at a distance of 14 min from the donor site) after inversion of the genomic region. Hence, the recognition of the zone for the reintegration of the phage DNA excised from malT is not affected by the inversion of a chromosomal region of about 26 min (between tnaA and purE) where it is located. In fact, after inversion, the prophage recognizes this region as though it were not inverted, since its reintegration occurs at a site which is only 3 min away from the integration site in the non-inverted chromosome.
The result changes dramatically if the donor site is in the invertible region. In this case, the excised prophage migrates to a new target, after inversion. In fact, the prophage migrates to about 57 min, before inversion, while, after inversion it goes to about 40 min. It is interesting to note that this last site is symmetrical with the preceding one, as was the localization of the lacY gene (where the prophage was originally inserted) before and after chromosome inversion.
Previous results, obtained with strain RS54, showed that the prophage excised from lacZ was reintegrated at a region around 74 min of the E. coli map. There is a difference with the results obtained with R430. In this case, the prophage was reintegrated around 57 min. We ascribe this difference to the great differences between the genetic background of RS54 and R340 rather than to the difference in the integration site of the prophage (lacZ in RS54 and lacY in R340). In R430 the donor site is within a genetic region which underwent significant rearrangements, due to the presence of two transposons and a transfer origin near the lac operon.
The replication-factory model predicts a symmetry between the two interacting sites as regards oriC. Most of the acceptor sites (4 out of 6) are symmetrical about oriC, and in the cases where donor and acceptor sites break down (the inversion strains) the replication symmetry might be altered because of ter sites. On the other hand, symmetry (or reciprocal results) may not be necessary in the case of the nucleoid-folding model and we have already observed that the Mu DNA excised from the gal operon (17 min) migrates to a chromosome region between 2 and 5 min, without any symmetry as regards oriC (Ghelardini et al., 1995 ). For these reasons, we believe that this last model could be more plausible. In any event, from these results, we can conclude that, during the excisionreintegration process, the receptor site is determined with regard to the location of the donor site.
Is the excisionreintegration process a general property of the Mu bacteriophage?
At this point, we may ask whether the excisionreintegration process is a peculiarity of the Mugem2ts mutant or a general property of the Mu bacteriophage. In fact, there is an important discrepancy between the behaviour of Mu wild-type and the Mugem2ts bacteriophages as far as the ability to give rise to precise excision is concerned.
Mugem2ts prophages are known to affect DNA supercoiling in host lysogens and to change the expression of many genes and a possible role of chromosome supercoiling in the ability of Mugem2ts to excise has been suggested (Ghelardini et al., 1994 ). The mutation was mapped within the gem promoter at position -44 (8153 bp from the left end of the phage genome) (La Valle et al., 1996
). Since the only genetic defect carried by Mugem2ts and involved in the aforesaid process can not be ascribed to a modified (thermosensitive) protein, but to modified expression of the wild-type gem operon, it should be possible to induce the excisionreintegration process of wild-type Mu in the presence of gem overexpression and, in particular, of GemA, which is responsible for the effects on DNA supercoiling. Preliminary results obtained with a recombinant plasmid carrying the gemA gene under the control of an inducible promoter indicate that in the presence of extra copies of the gemA gene, the wild-type Mu prophage is able to excise from its original localization and to reintegrate elsewhere on the host genome. Therefore, as far as the excisionreintegration process is concerned, it behaves as the gem2ts mutant (P. Ghelardini & J. C. Lièbart, unpublished). Hence, it is obvious that this process could be included in general Mu biology and could open up a new field of investigation on Mu illegitimate recombination, besides raising interesting questions on the nature of GemA activity.
![]() |
ACKNOWLEDGEMENTS |
---|
This work was partially supported by funds from the Ministero dell Università e della Ricerca Scientifica e Tecnologica and the Fondazione Cenci Bolognetti Istituto Pasteur (Italy).
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Casadeus, J. & Roth, J. R. (1989). Transcriptional occlusion of transposon targets.Mol Gen Genet 216, 204-209.[Medline]
Craig, N. L. (1997). Target site selection in transposition.Annu Rev Biochem 66, 437-474.[Medline]
Fuchs, R., Stoehr, P., Rice, P., Omond, R. & Cameron, G. (1990). New services of the EMBL data library.Nucleic Acids Res 18, 4319-4323.[Abstract]
Ghelardini, P., Lièbart, J. C., Di Zenzo, G., Micheli, G., DAri, R. & Paolozzi, L. (1994). A novel illegitimate recombination event: precise excision and reintegration with Mugem mutant prophage.Mol Microbiol 13, 709-718.[Medline]
Ghelardini, P., Lièbart, J. C., Fabozzi, G., Tomassini, B., DAri, R. & Paolozzi, L. (1995). Reversal of Mugem2ts-induced mutations.FEMS Microbiol Rev 17, 171-176.[Medline]
Grindley, N. D. & Reed, R. R. (1985). Transpositional recombination in prokaryotes.Annu Rev Biochem 54, 863-896.[Medline]
Kleckner, N. (1981). Transposable elements in prokaryotes.Annu Rev Genet 15, 341-404.[Medline]
Krug, P. J., Gileski, A. Z., Code, R. J., Torjussen, A. & Schmid, M. B. (1994). Endpoint bias in large Tn10-catalyzed inversions in Salmonella typhimurium. Genetics 136, 747-756.
La Valle, R., Fabozzi, G., Ghelardini, P. & Paolozzi, L. (1996). Bacteriophage Mugem2ts, which induces synchronous cell division in the infected host, is mutated in the gem operon promoter.Res Microbiol 147, 615-618.[Medline]
Lemon, K. & Grossman, A. D. (1998). Localization of bacterial DNA polymerase: evidence for a factory model of replication.Science 282, 1216-1219.
Mahillon, J. & Chandler, M. (1998). Insertion sequences.Microbiol Mol Biol Rev 62, 725-774.
Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory.
Mizobuchi, M. & Frohman, L. A. (1993). Rapid amplification of genomic DNA ends.Biotechniques 15, 213-216.
Pato, M. (1989). Bacteriophage Mu. In Mobile DNA, pp. 23-52. Edited by D. E. Berg & M. M. Howe. Washington, DC: American Society for Microbiology.
Rebollo, J. E., Francois, V. & Louarn, J. M. (1988). Detection and possible role of two large nondivisible zones on the Escherichia coli chromosome.Proc Natl Acad Sci USA 24, 9391-9395.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning; a Laboratory Manual, 2nd edn. Cold Spring Harbor, NewYork: Cold Spring Harbor Laboratory.
Shapiro, J. (1984). Observations on the formation of clones containing araBlacZ cistron fusions.Mol Gen Genet 194, 79-90.[Medline]
Wang, X. & Higgins, P. (1994). Muprints of the lac operon demonstrate physiological control over the randomness of in vivo transposition.Mol Microbiol 12, 665-677.[Medline]
Received 22 July 1999;
revised 1 November 1999;
accepted 6 December 1999.
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
J MED MICROBIOL | ALL SGM JOURNALS |