Comparative Studies with tox+ and tox? Corynebacteriophages

The characteristics of nine inducible temperate corynebacteriophages designated α^tox+, β^tox+, P^tox+, γ^tox−, π^tox+, K^tox−, ρ^tox−, L^tox+, and δ^tox+ have been compared. Virion morphology and ability to recombine genetically with the well-studied phage β^tox+ have been correlated with other properties of the phages, and the distribution of the genetic marker tox+ among related and relatively unrelated corynebacteriophages has been analyzed. The immunity specificity, host range, and plaque morphology of each phage were determined. The phages can be separated into five groups with different immunity specificities. Each type of host range previously recognized in mutants of phage β^tox+ was present in one or more of the phages included in the present study, and the phages were found to produce plaques of several different morphological types. Representative phages with each of the five types of immunity specificity were further characterized with respect to virion morphology, ability to recombine with phage β^tox+, latent period, average burst size, and neutralization by homologous and heterologous antiphage sera. All of these phages have polyhedral heads and long slender tails, but two distinct morphological types were distinguished by the sizes and proportions of the components of the virions. Only phages of the same morphological type as β^tox+ were capable of genetic recombination with β^tox+, but morphological similarity between phages was not sufficient to insure interfertility. The phages which recombined with β^tox+ resembled one another in plaque morphology, latent period, and average burst size, whereas phages which failed to recombine with β^tox+ differed in these characteristics. The phages capable of genetic recombination with β^tox+ were found to differ from each other in immunity specificity, host range, neutralization by antiphage sera, and toxinogenicity. Thus, these latter characteristics are of limited value in establishing the extent of relatedness between corynebacteriophages. The genetic marker tox+ was not consistently correlated with any other property of the corynebacteriophages analyzed in this study. The most striking finding regarding the distribution of the tox+ marker is its presence both in β^tox+ and δ^tox+, phages which fail to recombine genetically and which differ in virion morphology. The presence of the tox+ marker in genetically unrelated corynebacteriophages poses many questions concerning the origin(s) of tox+ and the evolution of the phage-host interactions which determine the ability of corynebacteria to synthesize diphtherial toxin.     

Genetic Analysis of tox+ and tox? Bacteriophages of Corynebacterium diphtheriae

A series of mutants derived from the temperate corynebacteriophages beta(tox+), gamma(tox-), and L(tox+) was isolated and characterized. In three-factor crosses between mutant beta phages the relative map order of the genetic markers determining extended host ranges (h and h') and loss of ability to lysogenize (c) was found to be h--c--h'. Recombination between markers was observed in matings between phage beta and the heteroimmune corynebacteriophages gamma and L. In such matings between heteroimmune phages the c markers of phages beta and gamma failed to segregate from the imm markers which determine the specificity of lysogenic immunity in these phages. The factor which directs the synthesis of diphtherial toxin during infection of appropriate corynebacterial hosts by toxinogenic corynebacteriophages is designated tox(+). It was possible to show that the tox(+) determinant of phage beta behaves as a single genetic element which occupies a position between the loci h and imm on the genetic map of this phage. Genetic recombination between mutants of phage beta occurred at very low frequencies in biparental matings performed by mixed infection of Corynebacterium diphtheriae C7(s)(-)(tox-). Considerably higher recombination frequencies were observed when lysogenic bacterial strains carrying one parental phage as prophage were induced by ultraviolet irradiation and then superinfected by the second parental phage. Maximal stimulation of genetic recombination between mutant beta phages was detected when superinfection followed ultraviolet irradiation of the lysogenic cells within a limited period of time. In matings between phages with incomplete genetic homology, the stimulation of recombination by ultraviolet radiation was much less effective.     

Integration of corynebacteriophages beta tox+, omega tox+, and gamma tox- into two attachment sites on the Corynebacterium diphtheriae chromosome.

The bacterial attachment sites of independently isolated Corynebacterium diphtheriae strains C7s and (belfanti)1030 lysogenic for corynebacteriophages beta tox+, omega tox+, and gamma tox- were determined by Southern blot analysis. Both corynebacterial strains contained two distinct bacterial attachment sites (attB1 and attB2). We found that infection by any of the three closely related corynebacteriophages may give rise to single, double, and triple lysogens. In the case of toxigenic C. diphtheriae strains C7s(beta tox+) and C7s(omega tox+), the final yields of diphtheria toxin produced under optimal conditions were equivalent and varied by one-, two-, or threefold depending upon the number of integrated prophage.     

Orientation of the tox gene in the prophage of corynebacteriophage beta.

The orientation of the gene for diphtheria toxin, tox, in the prophage of converting corynebacteriophage beta has been determined. The orientation of tox in prophage and that reported simultaneously by Holmes (1976) for vegetative phage are compatible with the hypothesis that beta phage is inserted into the chromosome of its bacterial host by means of a mechanism similar to that described for lambda phage, and that the phage attachment site lies between the tox and imm genes. The position of three tox mutations that are phenotypically CRM- has also been determined. Relative to the tox-45 mutation, they are located more proximally to the end of the tox structural gene that corresponds to the amino terminal of diphtheria toxin.     

Detection and physical map of a omega tox+-related defective prophage in Corynebacterium diphtheriae Belfanti 1030(-)tox-.

A library of chromosomal DNA from Corynebacterium diphtheriae Belfanti 1030(-)tox- was cloned in the lambda phage vector EMBL4 and screened for sequences homologous to corynephage omega tox+ and the attB1-attB2 region of the C7(-)tox- chromosome. Two portions of the 1030(-)tox- chromosome, 35 and 30.5 kilobases long which contain, respectively, the entire region homologous to corynephage omega tox+ and the attB1-attB2 sites, were mapped with the restriction endonucleases BamHI and EcoRI. Chromosomal DNA from 1030(-)tox- was shown to contain a 15.5-kilobase region that was homologous to ca. 42% of the corynephage omega tox+ genome. These sequences were found to hybridize to three regions of the phage genome and do not contain either the diphtheria tox operon or the attP site. These sequences are distant from the chromosomal region that contains the attB1-attB2 sites. Moreover, unlike other known defective prophages, the physical map of this prophage starts at the cos site and is colinear with the vegetative phage map. The 30.5-kilobase region of the 1030(-)tox- chromosome, which contains the attB1-attB2 sites, has a central core region that is almost identical to the corresponding region of the C7(-)tox- chromosome; however, the flanking sequences in these two strains of C. diphtheriae are different.     

Restriction endonuclease map of the nontoxigenic corynephage gamma c and its relationship to the toxigenic corynephage beta c.

Clear-plaque-forming mutant gamma tox- corynephages were isolated independently from nontoxigenic lysogenic Corynebacterium diphtheriae strains C7s(gamma tox-) and C4(gamma tox-). A physical map was constructed by using restriction endonucleases BamHI, EcoRI, HindIII, and KpnI. A comparison of nontoxigenic gamma c with toxigenic corynephage beta c revealed large areas of homology, including common regions for cohesive ends (cos) and attachment sites (att). Localization of the att sites on the beta prophage and correlation of the physical and genetic maps defined the orientation of the diphtheria tox operon. Diphtheria tox sequence homologies were mapped on gamma c by hybridizing 32P-labeled diphtheria tox mRNA to restriction fragments of gamma c DNA. Two regions of heterogeneity between phages beta c and gamma c were localized and these regions accounted for the 3-kilobase larger molecular size of gamma c compared with beta c. One change occurs near the tox promoter and may explain the nontoxigenic phenotype of corynephage gamma tox-.     

Comparative analysis of the DNA of Corynebacterium diphtheriae phages and the cloning of the gene determining diphtheria toxin synthesis

The analysis of the DNA of one nontoxigenic C. diphtheriae phage and two toxigenic ones has revealed that phage phi 984tox+ belongs to omega-like tox+ phages, phage phi 9tox+ is a representative of a new group of phages and phage B (Freeman) tox is a deletion mutant of phage beta. The location of this deletion on the physical map of this phage has been established. To obtain the physical map of phage phi 984tox+, the complete library of internal DNA fragments has been constructed in vector pBR 322. The gene of native diphtheria toxin has been cloned in vectors pBR 322 and pUR 250. Plasmids pUR 250 with the inserts of the toxin gene have been shown to be unstable if tox and lac promoters are located in tandem before the body of the toxin gene. The prolonged cultivation of clones having such structure leads to the formation of a spontaneous mutation located in the region coding the C-end part of the A-fragment of the toxin.     

Restriction endonuclease map of corynebacteriophage omega ctox+ isolated from the Park-Williams no. 8 strain of Corynebacterium diphtheriae.

The toxigenic corynebacteriophage omega tox+ was isolated from the hypertoxigenic Park-Williams no. 8 (PW8) strain of Corynebacterium diphtheriae and compared with the toxigenic corynebacteriophage beta tox+. The physical size and host range of both phages were found to be identical. An endonuclease restriction map of omega tox+ was constructed, and the locations of the cohesive ends (cos), phage attachment site (attP), and the diphtheria tox operon were identified. The genome of omega tox+ was found to differ from that of beta tox+ in three regions. In addition, omega tox+ was shown to be integrated into two nontandem corynebacterial phage attachment sites (attB1, attB2) in the PW8 chromosome. The differences in the restriction endonuclease digestion maps of omega tox+ and beta tox+ and the contribution of double lysogeny are discussed in relation to the hypertoxigenicity of the PW8 strain.     

Isolation and partial characterization of a corynebacteriophage beta, tox operator constitutive-like mutant lysogen of Corynebacterium diphtheriae.

We have isolated and partially characterized a beta-phage mutant lysogen of Corynebacterium diphtheriae, C7(betatoxct1+), which is partially insensitive to iron inhibition of diphtheria toxin production. tox expression by C7(betatoxct1+) was found to be partially constitutive. In the presence of concentrations of iron that almost completely inhibit the expression of diphtheria toxin by the wild type, C7(beta), the level of toxin production by C7(betatoxct1+) was found to be at least 25 times that of the parent. The purified tox gene product of C7(betatoxct1+) was immunologically and electrophoretically identical to, and equally as toxic as, diphtheria toxin purified from C7(beta). In addition, the partial N-terminal amino acid sequence was found to be identical to diphtheria toxin. This data strongly suggests that the mutation allowing for the constitutive expression of tox in C7(betatoxct1+) is outside of the structural gene. Furthermore, the constitutive expression of diphtheria toxin was found to be cis dominant in the double lysogen C7(betacrm45+/betatoxct1+). The data presented is consistent with the existence of a tox operator locus.     

Regulation of toxinogenesis in Corynebacterium diphtheriae. I. Mutations in bacteriophage beta that alter the effects of iron on toxin production

Diphtherial toxin is produced in maximal yields by Corynebacterium diphtheriae (C7(beta tox+) only when iron is present in growth-limiting amounts. Toxin production is markedly decreased under high-iron conditions. We studied the role of the bacteriophage beta genome in this apparent regulation of toxin production by iron. Using a passive immune hemolysis assay to detect toxin antigen production in individual plaques, we identified rare phage mutants that were toxinogenic in high-iron medium. Lysogenic derivatives of C. diphtheriae C7 harboring such phage mutants were constructed. The lysogens were compared with wild-type strain C7(beta) for their ability to produce toxin in deferrated liquid medium containing varying amounts of added iron. Quantitative tests for extracellular toxin were performed by competitive-binding radioimmunoassays. We identified phenotypically distinct mutant strains that produced slightly, moderately, or greatly increased yields of toxin antigen under high-iron conditions. The toxin produced by the mutant lysogens was biologically active and immunochemically indistinguishable from wild-type toxin. Complementation experiments demonstrated that the phage mutation designated tox-201 had a cis-dominant effect on the expression of the toxin structural gene of phage beta. The characteristics of the tox-201 mutation suggest that it defines a regulatory locus of phage beta that is involved in control of toxinogenesis by iron in C. diphtheriae.     

Isolation and characterization of tox mutants of corynebacteriophage beta.

Seventeen nontoxinogenic (tox) mutants of corynebacteriophage beta have been isolated by using a tissue culture screening technique. The mutants fall into four major classes. Two of the classes, I and II, appear to contain missense and nonsense mutants, respectively. However, classes III and IV have not been previously described. Class III mutants produce two proteins (CRMs) seriologically related to diphtheria toxin, but efforts to demonstrate the presence of more than one tox gene have been successful. Class IV mutants are phenotypically CRM-, failing to produce any detectable protein serologically related to diphtheria toxin. Genetic studies indicate that the mutations in class IV strains are not in a gene distinct form the structural gene for toxin, and that the CRM- strains retain at least a portion of that gene. A natural phage isolate, gamma, behaves in a completely parallel fashion to the class IV mutants. The production of tox+ recombinants through recombination of various pairs of tox phage mutants has been demonstrated. The implications of these findings for the natural history of diphtheria are discussed.     

Identification of deoxyribonucleic acid restriction fragments of beta-converting corynebacteriophages that carry the gene for diphtheria toxin.

Deoxyribonucleic acid fragments bearing the gene for diphtheria toxin have been identified in restriction enzyme digests of deoxyribonucleic acids from beta-converting and gamma-nonconverting corynebacteriophages. A combination of physical and genetic evidence has established that the Bam HI band C fragment of beta phage deoxyribonucleic acid, which carries the specific phage attachment site (Buck and Groman, J. Bacteriol. 148:131-142, 1981), also carries most, and probably all, of the gene for diphtheria toxin. A detailed restriction map of this tox-bearing Bam HI fragment has been developed, and the locations and orientation of the tox gene and the attP site within this fragment have been established.     

System for the Investigation of the Bacteriophage-directed Synthesis of Diphtherial Toxin

Lytic corynebacteriophage betahv64(tox+) has been characterized, and methods for studying the expression of its tox(+) gene in nontoxinogenic Corynebacterium diphtheriae strain C7(s)(-)(tox-) described. During one cycle of viral growth there was a 1 million-fold increase in extracellular toxin. Both the conditions of the experiment and the use of purified phage, free from toxin, support the conclusion that all of the toxin was newly formed. This toxin was immunochemically indistinguishable from standard toxin produced by the PW8(r)(Pdi)(tox+) strain. Chloramphenicol was found to be an effective agent for synchronizing the initiation of viral growth. Once chloramphenicol was removed, intracellular toxin appeared and continued to increase throughout the latent period. Proflavine, added early in the latent period, blocked phage maturation without similarly affecting yields of toxin. Iron exerted a limited inhibitory effect on final toxin levels attained.     

Differentiation of methicillin-resistant strains of Staphylococcus aureus by prophage specificity]

By inducing with mitomycin C the following phages were isolated from all the tested 32 methicillin resistant strains of S. aureus: the serogroup B phage was isolated from 2 strains, the serogroup B and F phages were isolated from 5 strains and the serogroup F phage was isolated from 25 strains. The phages were divided into 5 groups by the antiphage immunity. In group 1 of the phages 4 additional phages were specified. By the specificity of the prophages in the cultures all the strains were divided into 5 groups. Group 1 of the cultures was divided into 5 subgroups (A, B, C, D and E).     

DNA homology study of Corynebacterium diphtheriae v. gravis groups I, II and III, Corynebacterium ulcerans and Corynebacterium pseudotuberculosis (ovis)

The homology of genomes within Krylova 's groups I, II and III of C. diphtheriae, including toxigenic C. diphtheriae and their nontoxigenic precursors within the same group, was confirmed by the method of DNA/DNA molecular hybridization; the homology of DNA within the groups was 89-103%, the thermostability of heteroduplexes being high (on the level of homoduplexes ). The heterogeneity of genomes within these 3 groups of cultivar gravis was confirmed, which made it possible to consider C. diphtheriae, groups I, II and III, to belong to different, though closely related species; in intergroup hybridization the homology of DNA varied, as a rule, between 66% and 73%, while the thermostability of heteroduplexes was low: delta T50 was -3 degrees C to -6 degrees C. The differences in genomes (on the level of different species) between 3 groups of C. diptheriae v. gravis on one hand and C. diphtheriae v. mitis C7 (-) tox- and its convertant C7 (beta) tox+ of phage tox+ on the other hand (DNA homology being 56-62%), as well as between C. diphtheriae v. intermedius No. 328 tox+ on one hand and the representatives of 3 groups of C. diphtheriae v. gravis and C. diphtheriae v. mitis, strain C7 (beta) tox+, on the other hand (DNA homology being 42-43%) were revealed. The heterogeneity of genomes (on the level of different genera) was revealed between C. diphtheriae strains, cultivars gravis (groups I, II and III), mitis (C7(-) tox- and C7 (beta) tox+) and intermedius (No. 328 tox+) on one hand and C. ulcerans and C. pseudotuberculosis (ovis) strains on the other hand; DNA homology was 11-17% for C. ulcerans and 22-26% for C. pseudotuberculosis (ovis), the thermostability of heteroduplexes being at the lowest level (delta T50 was -11 degrees C to -13 degrees C). As a result, C. diphtheriae, classified by Bergey as a single species, was found to comprise 5 species detected by means of marking in accordance with their phenotypical features and genome structure, carried out by the method of DNA/DNA molecular hybridization; among these species were group I, II and III strains of cultivar gravis, strain C7 of cultivar mitis and strain No. 328 of cultivar intermedius. C. ulcerans and C. pseudotuberculosis (ovis) strains investigated in this study can possibly be placed outside the genus including 5 C. diphtheriae species.     

Physical mapping of beta-converting and gamma-nonconverting corynebacteriophage genomes

Deoxyribonucleic acids (DNAs) from wild-type and mutant strains of beta-converting and gamma-nonconverting corynebacteriophages were isolated and physically characterized. The data obtained from DNA heteroduplexes, restriction enzyme banding profiles, and restriction maps reinforce the conclusion that beta and gamma phages are very closely related. The major physical differences seen in the DNA heteroduplexes are a small substitution bubble and one or two insertions which are present on the gamma phage genome. The insertions account for the differences in the genome sizes of beta and gamma phages, and with the substitution they are responsible for most of the differences in the restriction endonuclease profiles and maps of the corynebactriophage genomes, two special sites and the DNA fragments carrying them were identified. These were the cohesive (cos) sites and the specific attachment (attP) site of the vegetative phage genome. The behavior of these sites indicated that the transition of phage DNA from the vegetative to the prophage state involves the circularization of vegetative DNA through the cos sites and its integration into the bacterial chromosome via the attP site. The mechanism of corynebacteriophage integration was similar to that employed by Escherichia coli phage gamma. From the data assembled the physical and genetic maps of beta and gamma phage were oriented with respect to one another. The extensive similarity in their maps provides additional confirmation of a close evolutionary relationship.     

Direct and general selection for lysogens of Escherichia coli by phage lambda recombinant clones.

We report a simple in vivo technique for introducing an antibiotic resistance marker into phage lambda. This technique could be used for direct selection of lysogens harboring recombinant phages from the Kohara lambda bank (a collection of ordered lambda clones carrying Escherichia coli DNA segments). The two-step method uses homologous recombination and lambda DNA packaging to replace the nonessential lambda DNA lying between the lysis genes and the right cohesive (cos) end with the neomycin phosphotransferase (npt) gene from Tn903. This occurs during lytic growth of the phage on a plasmid-containing host strain. Neomycin-resistant (npt+) recombinant phages are then selected from the lysates containing the progeny phage by transduction of a polA1 lambda lysogenic host strain to neomycin resistance. We have tested this method with two different Kohara lambda phage clones; in both cases, neomycin resistance cotransduced with the auxotrophic marker carried by the lambda clone, indicating complete genetic linkage. Linkage was verified by restriction mapping of purified DNA from a recombinant phage clone. We also demonstrate that insertion of the npt+ recombinant phages into the lambda prophage can be readily distinguished from insertion into bacterial chromosomal sequences.