DNA Rearrangements Associated with Reversion of Bacteriophage Mu-Induced Mutations

Excision of transposable genetic elements from host DNA is different from the classical prophage lambda type of excision in that it occurs at low frequency and is mostly imprecise; only a minority of excision events restores the wild-type host sequences. In bacteriophage Mu, a highly efficient transposon, imprecise excision is 10-100 times more frequent than precise excision. We have examined a large number of these excision events by starting with mucts X mutants located in the Z gene of the lac operon of Escherichia coli. Mucts X mutants are defective prophages whose excision occurs at a measurable frequency. Imprecise excision was monitored by selecting for melibiose+ (Mel+) phenotype, which requires only a functioning lacY gene. Mel+ revertants exhibit an array of DNA rearrangements and fall in four main classes, the predominant one being comprised of revertants that have no detectable Mu DNA. Most of these revertants can further revert to Lac+. Perhaps 5 base-pair duplications, originally present at prophage-host junctions, are left in these lacZ-Y+ revertants, and they can be further repaired to lacZ+. Another class has, in addition to the loss of Mu DNA, deletions that extend generally, but not always, to only one side of the prophage. The other two classes of revertants, surprisingly, still have Mu DNA in the lacZ gene. One class has deletions in the Z gene, whereas, no deletions can be detected in the other. Many of the revertants in the last class can further revert to lacZ+, indicating that the lacY gene must have been turned on by a rearrangement within Mu DNA. Apparently, all of the detectable precise and most of the imprecise excision events require functioning of the Mu A gene. We suggest that a block in large-scale Mu replication allows the excision process to proceed.     

A novel illegitimate recombination event: precise excision and reintegration with the Mu gem mutant prophage

The bacteriophage Mu is known to insert its DNA more or less randomly within the Escherichia coli chromosome, as do transposable elements, but unlike the latter, precise excision of the prophage, thereby restoring the original sequence, is not observed with wild-type Mu, although it has been reported with certain defective mutants. We show here that the mutant prophage Mu gem2ts can excise precisely from at least three separate loci —malT, Iac and thyA (selected as Mal⁺, Lac⁺ and Thy⁺, respectively). This excision occurs under permissive conditions for phage development, is observed in fully immune (c⁺) lysogens, and is independent of RecA and of Mu transposase. Mu gemts2 excision is invariably accompanied by reintegration of a Mu gem2ts prophage elsewhere in the chromosome, in the case of Mal⁺ revertants, this prophage is systematically located at 94min on the E. coli chromosome. Mu gem2ts excision therefore sheds some light on the long-standing paradox of the lack of precise Mu excisio.     

Transposable lambda placMu bacteriophages for creating lacZ operon fusions and kanamycin resistance insertions in Escherichia coli.

We have constructed several derivatives of bacteriophage lambda that translocate by using the transposition machinery of phage Mu (lambda placMu phages). Each phage carries the c end of Mu, containing the Mu cIts62, ner (cII), and A genes, and the terminal sequences from the Mu S end (beta end). These sequences contain the Mu attachment sites, and their orientation allows the lambda genome to be inserted into other chromosomes, resulting in a lambda prophage flanked by the Mu c and S sequences. These phages provide a means to isolate cells containing fusions of the lac operon to other genes in vivo in a single step. In lambda placMu50, the lacZ and lacY genes, lacking a promoter, were located adjacent to the Mu S sequence. Insertion of lambda placMu50 into a gene in the proper orientation created an operon fusion in which lacZ and lacY were expressed from the promoter of the target gene. We also introduced a gene, kan, which confers kanamycin resistance, into lambda placMu50 and lambda placMu1, an analogous phage for constructing lacZ protein fusions (Bremer et al., J. Bacteriol. 158:1084-1093, 1984). The kan gene, located between the cIII and ssb genes of lambda, permitted cells containing insertions of these phages to be selected independently of their Lac phenotype.     

Reversal of Mu gem2ts-induced mutations

Abstract: Mutations induced by the integration of a Mu gem 2ts mutant prophage can revert at frequencies around 1 × 10⁻⁶, more than 10⁴-fold higher than that obtained with Mu wild-type. Several aspects characterize Mu gem 2ts precise excision: (i) the phage transposase is not involved; (ii) the RecA protein is not necessary; and (iii) revertants remain lysogenic with the prophage inserted elsewhere in the host genome. In addition, prophage re-integration seems to be non-randomly distributed, whereas Mu insertion into the host genome is a transposition event without any sequence specificity. In this paper, we describe that the site of re-integration somehow depends on the original site of insertion. Two alternative models are proposed to explain the strong correlation between donor and receptor sites.     

The effect of gamma-irradiation on mu DNA transposition and gene expression

Mobile genetic elements are a ubiquitous presence in the genomes of all well-studied organisms. The effect of genomic stress on the status and transposition of these elements has not, as yet, been extensively characterized. We have been using temperate, transposable bacteriophage Mu as a model system to examine the behavior of mobile genetic elements and have previously shown that many DNA-damaging agents did not induce a Mu prophage to enter the lytic cycle of multiple rounds of DNA transposition. To extend these results and to examine the possibility that they were a reflection of damage to the DNA substrate for Mu transposition, we have constructed a mini-Mu plasmid, pMD12, which contains the early region of Mu, flanked by both extremities required for transposition in cis, and the beginning of the transposase gene A fused in frame to the lacZ gene. This A'-lacZ fusion protein maintains beta-galactosidase enzymatic activity under the control of the expression of the Mu transposase A gene and thus, the capacity for Mu transposition can be easily monitored by assaying for beta-galactosidase. By measuring the amount of beta-galactosidase after various doses of gamma-irradiation, we found that doses of up to 75 krad had no effect on the expression of the Mu transposase gene A. This was confirmed by the lack of induction of a Mu prophage in strains containing a chromosomally inserted Mu genome. Although the plaque-forming units per colony-forming unit of strain CSH67, containing a chromosomally inserted lambda prophage, increased approximately 100-fold from 0 to 75 krad, no stimulation of induction of prophage Mu lytic growth was observed. We also found that plasmid pMD12 did not transpose and chromosomally associate upon gamma-irradiation. This supports the assertion that DNA-damaging agents, including gamma-rays, do not induce the transposition of prokaryotic mobile genetic elements.     

A spontaneous Escherichia coli K12 mutant which inhibits the excision-reintegration process of Mu gem2ts

Escherichia coli K12 strains lysogenic for Mu gem2ts with the prophage inserted in a target gene (i.e., lacZ::Mu gem2ts lysogenic strains) revert to Lac⁺ by prophage precise excision with a relatively high frequency (about 1×10⁻⁶). The revertants obtained are still lysogens with the prophage inserted elsewhere in the bacterial chromosome. We have observed that, with the time of storage in stabs, bacterial cultures lysogenic for Mu gem2ts lose the ability to excise the prophage. The mutation responsible for this effect was co-transducible with the gyrB gene. After the removal of the prophage by P1 vir transduction from these strains, one randomly chosen clone, R3538, was further analyzed. It shows an increment of DNA supercoiling of plasmid pAT153, used as a reporter, and a reduced β-galactosidase activity. On the other hand, R3538 is totally permissive to both lytic and lysogenic cycles of bacteriophage Mu.     

Mu-lac insertion-directed mutagenesis in a pectate lyase gene of Erwinia chrysanthemi.

The pelC gene, which encodes one of the five major pectate lyase (PL) isoenzymes in Erwinia chrysanthemi 3937, designated PLc, was subcloned from a hybrid lambda phage into a pBR322 derivative and mutagenized with a mini-Mu-lacZ transposable element able to form fusions to the lacZ gene. One plasmid (pAD1) which had an inactivated pelC gene and a Lac+ phenotype was selected in Escherichia coli. This plasmid was introduced into Erwinia chrysanthemi, and the pelC::mini-Mu insertion was substituted for the chromosomal allele by homologous recombination. This strain lacks the PLc isoenzyme. This Erwinia chrysanthemi strain has a Lac+ phenotype that is inducible by polygalacturonate, as are the wild-type PL activities.     

The GemA protein of phage Mu and the GyrB gyrase subunit of Escherichia coli: the search for targets and interactions leading to the reversion of Mu-induced mutations

The mutant bacteriophage Mugem2(Ts), known to synchronize the division of infected cells, to relax DNA supercoiling and, as prophage, to give rise to precisely excised revertants, has been thought to overexpress the gemA-mor operon, and genetic evidence suggests that the B subunit of DNA gyrase (GyrB) is the target of action of GemA. In two different double hybrid tests presented here, we find no evidence of GemA-GyrB protein-protein interaction. We do observe a GemA-GemA interaction, however, indicating that GemA can dimerize. In lacZ::Mu lysogens, overexpression of the gemA-mor operon from a plasmid, under control of the L-arabinose inducible p(araBAD) promoter, does not permit the recovery of Lac(+) revertants. These observations suggest that GyrB is not the direct target of GemA action and that the various phenotypes of Mugem2(Ts) are not caused by overexpression of the gemA-mor operon.     

Conditionally transposition-defective derivative of Mu d1(Amp Lac).

A Mu d1 derivative is described which is useful for genetic manipulation of Mu-lac fusion insertions. A double mutant of the specialized transducing phage Mu d1(Amp Lac c62ts) was isolated which is conditionally defective in transposition ability. The Mu d1 derivative, designated Mu d1-8(Tpn[Am] Amp Lac c62ts), carries mutations which virtually eliminate transposition in strains lacking an amber suppressor. In such strains, the Mu d1-8 prophage behaves like a standard transposon. It can be moved from one strain of Salmonella typhimurium to another by the general transducing phage P22 with almost 100% inheritance of the donor insertion mutation. When introduced into a recipient carrying supD, supE, or supF, 89 to 94% of the Ampr transductants were transpositions of the donor Mu d1-8, from the transduced fragment into new sites. The stability of Mu d1-8 in a wild-type, suppressor-free background was sufficient to permit use of the fusion to select constitutive mutations without prior isolation of deletions to stabilize the fusion. Fusion strains could be grown at elevated temperature without induction of the Mu d prophage. The transposition defect of Mu d1-8 was corrected by a plasmid carrying the Mu A and B genes.     

Control of bacteriophage mu lysogenic repression

The transposable and temperate phage Mu infects Escherichia coli where it can enter the lytic life-cycle or reside as a repressed and integrated prophage. The repressor protein Rep is the key element in the lysis-lysogeny decision. We have analyzed the fate of Rep in different mutants by Western blotting under two conditions that can induce a lysogen: high temperature and stationary phase. We show that, unexpectedly, Rep accumulates under all conditions where the prophage is completely derepressed, and that this accumulation is ClpX-dependent. An analysis of the degradation kinetics shows that Rep is a target of two protease systems: inactivation of either the clpP or lon gene results in a stabilization of Rep. Such a reaction scheme explains the counterintuitive observation that derepression is correlated with high repressor concentration. We conclude that under all conditions of phage induction the repressor is sequestered in a non-active form. A quantitative simulation accounts for our experimental data. It provides a model that captures the essential features of Mu induction and explains some of the mechanisms by which the physiological signals affecting the lysis-lysogeny decision converge onto Rep.     

Model for the enchancement of lambde-gal integration into partially induced Mu-1 lysogens.

Temperate phage Mu-1, which is able to integrate at random in its host chromosome, is also able to mediate integration of other circular deoxyribonucleic acid, as a lambda-gal mutant unable to integrate by itself. After mixed infection with lambda-gal and Mucplus, galplus transductants are recovered that have the lambda-gal integrated in any circular permutation, sandwiched between two complete Mu genomes in the same orientation, the whole Mu-lambda-gal-Mu structure being found at any location in the bacterial chromosome. Here we show that such a lambda-gal can integrate in an induced Mu lysogen. In this case the lambda-gal is again in any circular permutation, between two Mu in the same orientation, but it is always located at the site of the original Mu prophage, and the two surrounding Mu have always the same genotype as the original Mu prophage. Active Mu replication functions are not essential for that process to occur. This suggests that bacterial replication may generate two Mu copies that in some way can regenerate a Mu attachment site that recombines with the lambda-gal. A model is presented that accounts for these observations, may be helpful for understanding some complex features of Mu development, and may possibly offer a basis for explaining spontaneous duplications.     

Genetic study of prophage excision with a temperature inducible mutant of Mu-1

Summary Escherichia coli strains, carrying F'episomes with a thermoinducible prophage of Mu were used to study the effect of heat induction on the ability of the cells to transfer these episomes to recipient cells. With this method information was obtained on the excision of the prophage after induction. Control studies were done with an episome containing a thermoinducible lambda prophage.The heat induction has opposite effects on the transfer of episomes containing Mu or : in the case of the number of sexductants in a -immune recipient is not significantly affected while sexduction into a non-immune recipient is increased by a factor 10–20. In contrast, the transfer of Mu-containing episomes into Mu-immune and also non-immune recipients is decreased by a factor of 100–200.The presented data exclude a precise -like excision of prophage Mu upon induction. A possible model for Mu excision is discussed.Colony forming units.     

Differential activity of a transposable element in Escherichia coli colonies.

In Escherichia coli colonies, patterns of differential gene expression can be visualized by the use of Mu d(lac) fusion elements. Here we report that patterned beta-galactosidase expression in colonies of strain MS1534 resulted from a novel mechanism, spatially localized replication of the Mu dII1681 element causing lacZ transposition to active expression sites. Mu dII1681 replication did not occur constitutively with a fixed probability but was dependent on the growth history of the bacterial population. The bacteria in which Mu dII1681 replication and lacZ transposition had occurred could no longer form colonies. These results lead to several interesting conclusions about cellular differentiation during colony development and the influence of bacterial growth history on gene expression and genetic change.     

Early events in the replication of Mu prophage DNA.

To determine whether the early replication of Mu prophage DNA proceeds beyond the termini of the prophage into hose DNA, the amounts of both Mu DNA and the prophage-adjacent host DNA sequences were measured using a DNA-DNA annealing assay after induction of the Mu vegetative cycle. Whereas Mu-specific DNA synthesis began 6 to 8 min after induction, no amplification of the adjacent DNA sequences was observed. These data suggest that early Mu-induced DNA synthesis is constrained within the boundaries of the Mu prophage. Since prophage Mu DNA does not undergo a prophage lambda-like excision from its original site after induction (E. Ljungquist and A. I. Bukhari, Proc. Natl. Acad. Sci. U.S.A. 74:3143--3147, 1977), we propose the existence of a control mechanism which excludes prophage-adjacent sequences from the initial mu prophage replication. The frequencies of the Mu prophage-adjacent DNA sequences, relative to other Escherichia coli genes, were not observed to change after the onset of Mu-specific DNA replication. This suggests that these regions remain associated with the host chromosome and continue to be replicated by the chromosomal replication fork. Therefore, we conclude that both the Mu prophage and adjacent host sequences are maintained in the host chromosome, rather than on an extrachromosomal form containing Mu and host DNA.     

Lactose metabolism in Erwinia chrysanthemi.

Wild-type strains of the phytopathogenic enterobacterium Erwinia chrysanthemi are unable to use lactose as a carbon source for growth although they possess a beta-galactosidase activity. Lactose-fermenting derivatives from some wild types, however, can be obtained spontaneously at a frequency of about 5 X 10(-7). All Lac+ derivatives isolated had acquired a constitutive lactose transport system and most contained an inducible beta-galactosidase. The transport system, product of the lmrT gene, mediates uptake of lactose in the Lac+ derivatives and also appears to be able to mediate uptake of melibiose, raffinose, and galactose. Two genes encoding beta-galactosidase enzymes were detected in E. chrysanthemi strains. That mainly expressed in the wild-type strains was the lacZ product. The other, the lacB product, is very weakly expressed in these strains. These enzymes showed different affinities for the substrates o-nitrophenyl-beta-D-galactopyranoside and lactose and for the inhibitors isopropyl-beta-D-thiogalactopyranoside and galactose. The lmrT and lacZ genes of E. chrysanthemi, together with the lacI gene coding for the regulatory protein controlling lacZ expression, were cloned by using an RP4::miniMu vector. When these plasmids were transferred into Lac- Escherichia coli strains, their expression was similar to that in E. chrysanthemi. The cloning of the lmrT gene alone suggested that the lacZ or lacB gene is not linked to the lmrT gene on the E. chrysanthemi chromosome. One Lac+ E. chrysanthemi derivative showed a constitutive synthesis of the beta-galactosidase encoded by the lacB gene. This mutation was dominant toward the lacI lacZ cloned genes. Besides these mutations affecting the regulation of the lmrT or lacB gene, the isolation of structural mutants unable to grow on lactose was achieved by mutagenic treatment. These mutants showed no expression of the lactose transport system, the lmrT mutants, or the mainly expressed beta-galactosidase, lacZ mutants. The lacZ mutants retained a very low beta-galactosidase level, due to the lacB product, but this level was low enough to permit use of the lacZ mutants for the construction of gene fusions with the Escherichia coli lac genes.     

Somatic excision of the Mu1 transposable element of maize.

The Mu transposons of the Robertsons's Mutator transposable element system in maize are unusual in many respects, when compared to the other known plant transposon systems. The excision of these elements occurs late in somatic tissues and very rarely in the germ line. Unlike the other plant transposons, there is no experimental evidence directly linking Mu element excision and integration. We have analyzed the excision products generated by a Mu1 transposon inserted into the bronze 1 locus of maize. We find that the excision products or 'footprints' left by the Mu1 element resemble those of the other plant transposable elements, rather than those of the animal transposable element systems. We also find some novel types of footprints resembling recombinational events. We suggest that the Mu1 element can promote intrachromosomal crossovers and conversions near its site of insertion, and that this may be another mechanism by which transposons can accelerate the evolution of genomes.     

Effect of mutations of Escherichia coli K-12 chromosome on the integration of bacteriophage Mu introduced into cells by infection or conjugation

We present the detailed research on the previously described Escherichia coli K-12 Mud- mutants with impaired development of bacteriophage Mu. The ability of Mu phage DNA to penetrate into mutant cells on infection was shown. If introduced into the cells or combined with mud mutation by recombination, the prophage may be induced, which results in phage Mu lythic development and phage burst from mutant cells. In the course of conjugative transfer into the mutant cells, within a DNA fragment of the lysogenic donor chromosome, MupAp1 prophage is not inherited by recombinants. At the same time, Mu prophage deficient in genes A and B, whose products are required for transposition, is inherited by the mutant with the usual frequency. These data enable us to conclude that the mud mutations disturb the stage of conservative transposition which is connected with the insertion of the Mu prophage into the chromosome, after excision from the linear DNA introduced into the cells via infection or conjugation.     

Bacteriophage mu-induced deletions in a plasmid containing the nif (N2 fixation) genes of Klebsiella pneumoniae.

Five plasmids with insertions of a heat-inducible Mu prophage in a Mu-sensitive and P1-sensitive derivative of plasmid pRD1, a recombinant R factor containing the his-nif region of Klebsiella pneumoniae, were isolated and characterized. In one plasmid containing the Mu prophage integrated at the his-distal end of nif, selection for heat resistance resulted in the generation of deletions extending from the Mu prophage into the nif region. Thirty of these deltions were used to map 26 point mutations in nif.     

Analysis of genetic recombination between two partially deleted lactose operons of Escherichia coli K-12.

Genetic recombination between a nontandem duplication of two partially deleted lactose operons (lacMS286phi80dIIlacBK1) in Escherichia coli K-12 has been examined. Since the deletions were nonoverlapping, rare lactose-fermenting (Lac+) recombinants occurred and were detected qualitatively on lactose tetrazolium agar indicator plates as white papillae growing on the surface of red colonies or quantitively on lactose minimal agar plates. Formation of Lac+ recombinants required the recA, recB, and recC gene products. Indirect suppression of recB21 by sbcB15 led to an increase in the frequency of Lac+ recombinants over wild-type levels. recF143 did not appreciably alter the number of Lac+ progeny, whereas recL152 and sbcB15 strains yielded increased numbers of Lac+ recombinants. The nature and formation of Lac+ recombinants was also examined. Respreading analysis indicated that formation of recombinants occurred primarily as the cells entered early stationary phase on the surface of the minimal agar plates and that over 90% of the recombinants contained a phi80dIIlac+ prophage. Time-of-entry experiments suggested that the region of deoxyribonucleic acid between the two operons was not inverted as a result of the recombinational event.