Vol. 172, No. 2
JOURNAL OF BACTERIOLOGY, Feb. 1990, p. 932-941 0021-9193/90/020932-10$02.00/0 Copyright © 1990, American Society for Microbiology
Bacteriophage Mu as a Genetic Tool To Study Erwinia amylovora Pathogenicity and Hypersensitive Reaction on Tobacco JOEL L. VANNESTE,l.2* JEAN-PIERRE PAULIN,2 AND DOMINIQUE EXPERT1 Pathologie Vgetale, 75231 Paris Cedex 05,1 and Institut National de la Recherche Agronomique Station de Pathologie Vegetale, Route de St Clement, Beaucouze 49000 Angers,2* France Received 5 June 1989/Accepted 17 October 1989
Erwinia amylovora 1430 was shown to be sensitive to Mu G(-) particles. Infection resulted either in lytic development or in lysogenic derivatives with insertion of the Mu genome at many sites in the bacterial chromosome. We used the Mu dl BX::Tn9 (lac Apr Cmr) derivative, called Mu dX, to identify mutans aected in pathogenicity and in their ability to induce a hypersensitive reaction (HR) on tobacco plants. Inoculation of 1,400 lysogenic derivatives on apple root calli led to the identification of 12 mutants in three classes: (i) class on tobacco plants; (ii) class 2 mutants were 1 mutants were nonpathogenic and unable to induce an nonpathogenic but retained the ability to induce an HR; and (iii) class 3 mutants showed attenuated virulence. Of the 12 mutants, 8 had a singe insertion of the Mu dX prophage. For class 1 and 2 mutants, reversion to pathogenicity was concomitant with the loss of the Mu dX prophage. Furthermore, revertants from the class 1 mutants also recovered the ability to induce an HR on tobacco plants. Five of the six class 3 mutants were impaired in exopolysaccharide production. No changes of the envelope structure (lipopolysaccharide and outer membrane proteins) were correlated with differences in pathogenicity. One class 3 mutant did not produce any functional siderophore, suggesting that iron uptake could be involved in pathogenicity. Erwinia amylovora causes fire blight, a typical necrotic disease which affects all plant species of the Pomoideae, but is especially destructive to pear and apple trees. Symptoms include water soaking, wilting, and rapid necrosis of succulent tissues (59). Under favorable climatic conditions, drops of ooze consisting of bacteria embedded in exopolysaccharides (EPS) are produced from infected tissues. In vitro, E. amylovora produces these same EPS, which are either tightly bound to the cell to form a capsule or run freely on the medium to constitute the slime (2, 7, 53). To date, EPS are the only components known to be involved in the pathogenicity of this bacterium (5, 6); however their role is still controversial (4, 25, 27, 31, 46, 52). Recently, Steinberger and Beer isolated TnS-induced mutants of E. amylovora affected in pathogenicity by using the suicide vector pJB4JI (55). Unfortunately, this plasmid is stable in E. amylovora (55; J. L. Vanneste, unpublished data). The authors further found that most of the mutants carried more than one insertion element, such as the Mu prophage transposed from pJB4JI or several copies of TnS and IS50. These multiple insertions complicated the genetic analysis of the mutants obtained. Such problems with pJB4JI have previously been encountered in other bacterial species (41). Bacteriophage Mu and its derivatives have been used for a variety of genetic manipulations, including transposon mutagenesis in soft-rot erwinias (1, 22). In this paper, we show that E. amylovora 1430 is sensitive to bacteriophage Mu and that Mu derivatives can be used to generate insertions at many sites in the bacterial chromosome. Another factor that conditions the success in identifying pathogenicity determinants is the availability of a reliable specific pathogenicity test that is rapid and simple enough to allow the screening of a large number of bacterial clones. We used apple root calli to screen 1,400 lysogenic clones of E. amylovora obtained by using the Mu derivative Mu dl BX::Tn9, also called Mu dX (3). Twelve mutants affected in pathogenicity or in virulence *
were identified and characterized for their ability to induce a hypersensitive reaction (HR) when injected in nonhost tissues. The envelope structure of the mutants, including outer membrane proteins and lipopolysaccharide (LPS), as well as their ability to grow on medium containing an iron
chelator, ethylenediamine-N,N'-bis(2-hydroxyphenylacetic acid) (EDDA), were also studied. MATERIALS AND METHODS Strains and media. The bacterial strains and bacteriophages used in this study are described in Table 1. Except when otherwise specified, E. amylovora strains were incubated at 27°C, and Escherichia coli strains were incubated at 37°C or at 30°C when they contained a Mu cts62 prophage. Complete medium was Luria broth (L broth) (40). Minimal medium was M9 (40); it contained 2 g of glucose per liter as the carbon source and was supplemented with nicotinic acid (50 mg/liter) to support the growth of E. amylovora. When production of EPS was determined, strains were grown on YP medium (6). Iron-deficient medium was obtained by the addition of 100 ,ug of EDDA (Fluka AG Chemical Co., Neu Ulm, Federal Republic of Germany) per ml. EDDA stock solutions were deferrated by the procedure of Rogers (45). When required, the media were solidified by using agar (12 g/liter for plates and 7 g/liter for overlays; Difco Laboratories, Detroit, Mich.). When needed, supplements were added at the following concentrations: chloramphenicol, 20 mg/liter; amino acids, purines, or pyrimidines, 50 to 200 jig/liter. Auxotrophic requirements were characterized by the nutritional pool system described by Davis et al. (19). Preparation of bacteriocin solutions and bacteriocin sensitivity tests. Bacteriocin synthesis was induced by adding mitomycin C to a culture grown in L broth as described by Expert and Toussaint (23). Sensitivity to bacteriocin was determined by spotting drops of various dilutions of the bacteriocin solution onto an overlay seeded with the culture to be tested. E. amylovora 1430 was sensitive to the bacteriocins produced by strains 651 and 5778 and partially
Corresponding author. 932
NONPATHOGENIC Mu-INDUCED MUTANTS OF E. AMYLOVORA
VOL. 172, 1990
933
TABLE 1. Bacterial strains, bacteriophages, and plasmid E. amylovora 1430 JLV 522, JLV 523, JLV 524, JLV 525, JLV 526 JLV 631, JLV 632, JLV 633
Source or reference
Characteristics
Strain, phage, or plasmid
Wild type, isolated from Crataegus sp. Mu dX lysogenic derivatives of 1430, auxotrophic
44 This work
Mu dX lysogenic derivatives of 1430, nonpathogenic and HR-a Mu dX lysogenic derivatives of 1430, nonpathogenic Mu dX lysogenic derivatives of 1430, affected in virulence
This work
F' pro lacZ8305::Mu cts62/Mu dX (pro lac) hsdR514 supE44 supFS8 lacYl galK2 galT22 metBI trpR55; K-12 derivative sensitive to Mu G(+) particles
3 29 (from A. Toussaint)
Wild type, sensitive to Mu G(-) Wild-type bacteriocin-producing strains
26 23
20D3 20D31 5778
Wild-type bacteriocin-producing strain Wild-type bacteriocin-producing strain Wild-type bacteriocin-producing strain
ATCC 19321 (from M. Faelen) NCPPB 802 (from M. Faelen) NCPPB 577-8 (from M. Faelen)
E. herbicola 651
Wild-type bacteriocin-producing strain
Laboratoire Botanique, University of Liege (from M. Faelen)
Bacteriophages Mu cts62 Mu cts62 S::Tn9 Mu cts62 7701
Thermoinducible Produces only G(-) particles Produces only G(+) particles Mu dl BX::Tn9 (lac Apr Cm')
30 13 Thompson and Howe (unpublished) (from A. Toussaint) 3
Apr Tcr Cmr
10
JLV 634, JLV 635, JLV 636 JLV 637, JLV 638, JLV 639, JLV640, JLV641, JLV 642 E. coli CAG 5050 BHB 2600
E. chrysanthemi B374 1277, 1456, 1500, 1521, 1871, 1884
This work This work
E. uredovora
Mu dX
Plasmid pBR325 a
HR- does not induce hypersensitivity on tobacco plant.
sensitive (as shown by the presence of a turbid zone) to the bacteriocins produced by strains 20D3 and 20D31. No inhibition was observed with undiluted bacteriocin solutions from E. chrysanthemi 1277, 1456, 1500, 1521, 1871, or 1884. Sensitivity to this set of bacteriocins was used to confirm the identity of strain 1430. Preparation of bacteriophage lysates. Mu bacteriophage lysates were prepared from lysogenic strains of E. coli by thermal induction as described by Toussaint and Schoonejans (57). Plate lysates were made in SM buffer (61), as described by Toussaint and Schoonejans (57). Suspensions were sterilized by addition of a few drops of chloroform and centrifuged to remove cell debris. The titer of Mu stocks was determined by spotting 10-pul drops of 1/10 serial dilutions onto overlays seeded with either E. chrysanthemi B374 or E. coli BHB 2600 and incubated at 35 and 42°C, respectively. Isolation of lysogenic strains of E. amylovora. Mu infection was performed as described by Baker et al. (3). The strain of E. amylovora to be infected was grown to 2 x 108 cells per ml in L broth supplemented with 2.5 mM CaCl2 and 5 mM MgSO4. A fresh phage lysate was added to obtain a multiplicity of infection of the helper phage of between 0.1 and 0.05. Phages were allowed to adsorb at room temperature for 30 min. The infected cells were then diluted 1:10 with L
broth and incubated with shaking at 27°C for 30 min to allow phenotypic expression. Selection of lysogenic clones of strain 1430, after infection with a Mu dX lysate, was based on the chloramphenicol resistance conferred by the Tn9 carried by the Mu dX prophage. The frequency of lysogenization was calculated as the number of bacteria that acquired the antibiotic resistance divided by the total number of bacteria infected. Identification of mutants affected in pathogenicity. The screening was done by individual inoculation of apple root calli. The calli were propagated as described by Paulin and Duron (43). Calli, 3 to 4 weeks old, were cut into quarters just before inoculation. The strains to be tested were transferred from a 24-h plate with a toothpick and applied to the freshly made cut. Calli were incubated in the dark at 17°C. After 3 days of incubation, when exudate produced by the wild-type strain could already be seen, the calli were examined daily for the presence of bacterial exudate under x6 magnification. When no exudate was detected after 13 days of incubation, the strain was retested at least two more times. Pathogenicity and virulence tests. Pathogenicity tests were done on young plants of Malus domestica cv. Golden Delicious, Pyrus communis cv. Passe Crassane, and Pyra-
934
VANNESTE ET AL.
cantha cv. Rogersiana. Inocula were prepared by suspending bacteria, grown for 12 to 16 h on L agar plates, in sterile water to a final concentration of about 5 x 108 CFU/ml. We inoculated plants by cutting through the main vein of a young, fully developed leaf with scissors previously dipped in the bacterial suspension. Routinely, three plants with two leaves inoculated per plant were used per strain. When virulence was quantified, a 10-,ul drop of the bacterial suspension was dispensed between two parallel cuts made on the main vein of a young, fully developed leaf. This test was done on 10 plants per strain and repeated at least three times. Sterile distilled water was used as a control. Plants were maintained in a growth chamber at 21 to 24°C under 10,000 lux with a 12-h light, 12-h dark cycle. Under these conditions the wild-type strain 1430 produced typical symptoms within 2 days, followed by necrosis of the infected tissues. When immature pear fruits were used for the pathogenicity test, we followed the procedure described by Steinberger and Beer (55). HR on tobacco. HR-inducing ability was tested on leaves of the tobacco cultivar Xanthi. Inocula were prepared in sterile distilled water as described for the pathogenicity test. Infiltration of the intercostal area of young and fully expanded leaves was done with 18-gauge sterile needles. Sterile water was used as a control. Routinely, three infiltrations were done per strain. The tobacco plants were kept in a growth chamber under the same conditions as described for the pathogenicity and virulence tests. Under these conditions the wild-type strain 1430 developed a typical HR within 12 h. Isolation of virulent revertants. Virulent revertants of the mutants of classes 1 and 2 were recovered from drops of exudate produced on immature pear fruits inoculated with high bacterial populations. Immature pear fruits of cultivar Bartlett kept in the dark at 0 to 4°C with moderate aeration were surface disinfected with 70% ethanol. Slices were made and placed in plastic boxes on sterile paper towels previously moistened with sterile water. L broth (20 ml) was inoculated with a single colony of the mutant to be analyzed. Overnight cultures were centrifuged, and the pellet was suspended in 10 ml of 10 mM MgSO4 solution. Then 100 to 200 ,ul of this bacterial suspension was applied per slice. As soon as the bacterial suspension was absorbed, the fruits were incubated in the dark at 27°C for 12 days. We inoculated 40 to 70 pear slices per mutant. As a control, 50 pear slices were each inoculated with 100 to 200 ixl of 10 mM MgSO4 solution. After 2 weeks of incubation no exudate was detected. When a drop of ooze was detected on a pear slice inoculated with a mutant, it was immediately streaked out onto L medium plates. Twenty colonies per isolate were randomly chosen to determine the sensitivity to chloramphenicol. One colony resistant to chloramphenicol and one sensitive to chloramphenicol were used for pathogenicity and HR tests. DNA methods. Total DNA from E. amylovora was isolated as described by Klotz and Zimm (35) and purified by ethidium bromide-cesium chloride centrifugation (39). Plasmid DNA was extracted by using the procedure of Kado and Liu (33). When used as probe, the plasmid DNA was isolated from overnight cultures by a cleared-lysate procedure (15) and further purified by ethidium bromide-cesium chloride centrifugation (39). Mu DNA isolated from a lysogenic strain of Proteus mirabilis was a generous gift from A. Toussaint and M. Faelen, Universite Libre de Bruxelles, Brussels, Belgium. Restriction endonucleases were purchased from Bethesda Research Laboratories, Inc., Gaithersburg, Md.,
J. BACTERIOL.
and used as specified by the supplier. Phage and plasmid DNA were labeled by using a nick translation reagent kit purchased from Amersham Corp., Arlington Heights, Ill., with 100 LCi of [a-32P]dCTP used per reaction. The conditions for DNA labeling were those specified by the supplier. DNA electrophoresis was done as described by Maniatis et al. (39). The membranes used to transfer the DNA were either Biodyne 3G (Pall Biodyne) or GeneScreen Plus (Du Pont, NEN Research Products, Boston, Mass.). When Biodyne 3G membranes were used, DNA transfer and hybridization were carried out as described by Silhavy et al. (50). When using GeneScreen Plus we followed the instructions given by the manufacturer. Autoradiography was carried out at -80°C with XAR-5 film (Eastman Kodak Co., Rochester, N.Y.) Cronex Lightning-Plus intensifying screens (E. I. du Pont de Nemours & Co., Wilmington, Del.). Capsule staining. Capsule and slime layers were revealed by the wet India ink staining procedure of Duguid as described by Doetsch (20), with the following modifications. The strains to be tested were grown for 24 h on a carbohydrate-rich medium (YP medium). A colony was suspended onto a glass slide in a drop of sterile water and thoroughly mixed with a drop of India ink (Pelikan). Outer membrane preparations and analysis. Triton-insoluble walls (outer membranes) were prepared and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described by Expert and Toussaint (23). The proteins used as standards to calculate the apparent molecular sizes were purchased from Bio-Rad Laboratories, Richmond, Calif. LPS extraction and analysis. LPS was extracted from E. amylovora by using the miniscale procedure designed by Schoonejans et al. (48). Samples were analyzed by SDSPAGE on 13.5% acrylamide gels and stained by the silverstaining method of Hitchcock and Brown (28). Detection of a siderophore activity. Siderophore activity was detected by the chemical assay devised by Schwyn and Neilands (49). When plated onto the blue medium supplemented with nicotinic acid, colonies of strain 1430 were surrounded by an orange halo, indicating the production and excretion of a siderophore. RESULTS Strain 1430 is sensitive to phage Mu. The Mu genome contains a region, called G, that can invert. Each orientation of this DNA fragment corresponds to the synthesis of different proteins involved in the host specificity of the viral particle. The two different kinds of particles are called Mu G(+) and Mu G(-) according to the orientation of the G fragment in the Mu DNA (58). When a Mu cts62 lysate, containing both Mu G(+) and Mu G(-) particles, was spotted onto an overlay seeded with E. amylovora 1430, zones of partial inhibition arose within 16 h of incubation at 35°C. Mu lysates grown by infection of strain 1430 were made from these inhibition zones. The titer of these lysates was 1,000 times higher on E. chrysanthemi B374, which is sensitive to Mu G(-) particles, than on E. coli BHB 2600, which is sensitive to Mu G(+) particles. Since BHB 2600 is restriction negative and B374 was shown not to restrict the growth of Mu phages originating from different host species (57), the difference in plaque production observed between B374 and BHB 2600 suggests that strain 1430 is sensitive to Mu G(-) particles. Further evidence was sought by using the Mu cts62 S::Tn9 derivative that can produce only Mu G(-) particles and the Mu cts62 7701 derivative that can
NONPATHOGENIC Mu-INDUCED MUTANTS OF E. AMYLOVORA
VOL. 172, 1990
TABLE 2. Frequency of reversion to wild-type phenotype of auxotrophic Mu dX lysogenic derivatives of strain 1430 and sensitivity of the revertants to chloramphenicol and to Mu Mutant strain
JLV JLV JLV JLV JLV
522 523 524 525 526
Frequency of Axtoh
reesnto
Auxtotrophy thrvrint rototrophyof paA
Adenine Threonine Arginine Methionine Tryptophan
S 8 8 8 4
x
10-9
x 10-10 x 10-10 x 10-9 x
10-8
No. of
Cm' strains/no, tested
20/20 8/8 6/6 12/12 20/20
No. of Mu' strains/no. tested
5/5 2/2 2/2 3/3 NDa
a ND, Not determined.
produce only Mu G(+) particles. Although no growth on strain 1430 of either of these derivatives was detected, chloramphenicol-resistant derivatives of strain 1430 were recovered at a frequency of 7 x 10-6 when Mu cts62 S: :Tn9 lysates were used, whereas no lysogenic clones could be isolated when using Mu cts62 7701 lysates (frequency,
JOURNAL OF BACTERIOLOGY, Feb. 1990, p. 932-941 0021-9193/90/020932-10$02.00/0 Copyright © 1990, American Society for Microbiology
Bacteriophage Mu as a Genetic Tool To Study Erwinia amylovora Pathogenicity and Hypersensitive Reaction on Tobacco JOEL L. VANNESTE,l.2* JEAN-PIERRE PAULIN,2 AND DOMINIQUE EXPERT1 Pathologie Vgetale, 75231 Paris Cedex 05,1 and Institut National de la Recherche Agronomique Station de Pathologie Vegetale, Route de St Clement, Beaucouze 49000 Angers,2* France Received 5 June 1989/Accepted 17 October 1989
Erwinia amylovora 1430 was shown to be sensitive to Mu G(-) particles. Infection resulted either in lytic development or in lysogenic derivatives with insertion of the Mu genome at many sites in the bacterial chromosome. We used the Mu dl BX::Tn9 (lac Apr Cmr) derivative, called Mu dX, to identify mutans aected in pathogenicity and in their ability to induce a hypersensitive reaction (HR) on tobacco plants. Inoculation of 1,400 lysogenic derivatives on apple root calli led to the identification of 12 mutants in three classes: (i) class on tobacco plants; (ii) class 2 mutants were 1 mutants were nonpathogenic and unable to induce an nonpathogenic but retained the ability to induce an HR; and (iii) class 3 mutants showed attenuated virulence. Of the 12 mutants, 8 had a singe insertion of the Mu dX prophage. For class 1 and 2 mutants, reversion to pathogenicity was concomitant with the loss of the Mu dX prophage. Furthermore, revertants from the class 1 mutants also recovered the ability to induce an HR on tobacco plants. Five of the six class 3 mutants were impaired in exopolysaccharide production. No changes of the envelope structure (lipopolysaccharide and outer membrane proteins) were correlated with differences in pathogenicity. One class 3 mutant did not produce any functional siderophore, suggesting that iron uptake could be involved in pathogenicity. Erwinia amylovora causes fire blight, a typical necrotic disease which affects all plant species of the Pomoideae, but is especially destructive to pear and apple trees. Symptoms include water soaking, wilting, and rapid necrosis of succulent tissues (59). Under favorable climatic conditions, drops of ooze consisting of bacteria embedded in exopolysaccharides (EPS) are produced from infected tissues. In vitro, E. amylovora produces these same EPS, which are either tightly bound to the cell to form a capsule or run freely on the medium to constitute the slime (2, 7, 53). To date, EPS are the only components known to be involved in the pathogenicity of this bacterium (5, 6); however their role is still controversial (4, 25, 27, 31, 46, 52). Recently, Steinberger and Beer isolated TnS-induced mutants of E. amylovora affected in pathogenicity by using the suicide vector pJB4JI (55). Unfortunately, this plasmid is stable in E. amylovora (55; J. L. Vanneste, unpublished data). The authors further found that most of the mutants carried more than one insertion element, such as the Mu prophage transposed from pJB4JI or several copies of TnS and IS50. These multiple insertions complicated the genetic analysis of the mutants obtained. Such problems with pJB4JI have previously been encountered in other bacterial species (41). Bacteriophage Mu and its derivatives have been used for a variety of genetic manipulations, including transposon mutagenesis in soft-rot erwinias (1, 22). In this paper, we show that E. amylovora 1430 is sensitive to bacteriophage Mu and that Mu derivatives can be used to generate insertions at many sites in the bacterial chromosome. Another factor that conditions the success in identifying pathogenicity determinants is the availability of a reliable specific pathogenicity test that is rapid and simple enough to allow the screening of a large number of bacterial clones. We used apple root calli to screen 1,400 lysogenic clones of E. amylovora obtained by using the Mu derivative Mu dl BX::Tn9, also called Mu dX (3). Twelve mutants affected in pathogenicity or in virulence *
were identified and characterized for their ability to induce a hypersensitive reaction (HR) when injected in nonhost tissues. The envelope structure of the mutants, including outer membrane proteins and lipopolysaccharide (LPS), as well as their ability to grow on medium containing an iron
chelator, ethylenediamine-N,N'-bis(2-hydroxyphenylacetic acid) (EDDA), were also studied. MATERIALS AND METHODS Strains and media. The bacterial strains and bacteriophages used in this study are described in Table 1. Except when otherwise specified, E. amylovora strains were incubated at 27°C, and Escherichia coli strains were incubated at 37°C or at 30°C when they contained a Mu cts62 prophage. Complete medium was Luria broth (L broth) (40). Minimal medium was M9 (40); it contained 2 g of glucose per liter as the carbon source and was supplemented with nicotinic acid (50 mg/liter) to support the growth of E. amylovora. When production of EPS was determined, strains were grown on YP medium (6). Iron-deficient medium was obtained by the addition of 100 ,ug of EDDA (Fluka AG Chemical Co., Neu Ulm, Federal Republic of Germany) per ml. EDDA stock solutions were deferrated by the procedure of Rogers (45). When required, the media were solidified by using agar (12 g/liter for plates and 7 g/liter for overlays; Difco Laboratories, Detroit, Mich.). When needed, supplements were added at the following concentrations: chloramphenicol, 20 mg/liter; amino acids, purines, or pyrimidines, 50 to 200 jig/liter. Auxotrophic requirements were characterized by the nutritional pool system described by Davis et al. (19). Preparation of bacteriocin solutions and bacteriocin sensitivity tests. Bacteriocin synthesis was induced by adding mitomycin C to a culture grown in L broth as described by Expert and Toussaint (23). Sensitivity to bacteriocin was determined by spotting drops of various dilutions of the bacteriocin solution onto an overlay seeded with the culture to be tested. E. amylovora 1430 was sensitive to the bacteriocins produced by strains 651 and 5778 and partially
Corresponding author. 932
NONPATHOGENIC Mu-INDUCED MUTANTS OF E. AMYLOVORA
VOL. 172, 1990
933
TABLE 1. Bacterial strains, bacteriophages, and plasmid E. amylovora 1430 JLV 522, JLV 523, JLV 524, JLV 525, JLV 526 JLV 631, JLV 632, JLV 633
Source or reference
Characteristics
Strain, phage, or plasmid
Wild type, isolated from Crataegus sp. Mu dX lysogenic derivatives of 1430, auxotrophic
44 This work
Mu dX lysogenic derivatives of 1430, nonpathogenic and HR-a Mu dX lysogenic derivatives of 1430, nonpathogenic Mu dX lysogenic derivatives of 1430, affected in virulence
This work
F' pro lacZ8305::Mu cts62/Mu dX (pro lac) hsdR514 supE44 supFS8 lacYl galK2 galT22 metBI trpR55; K-12 derivative sensitive to Mu G(+) particles
3 29 (from A. Toussaint)
Wild type, sensitive to Mu G(-) Wild-type bacteriocin-producing strains
26 23
20D3 20D31 5778
Wild-type bacteriocin-producing strain Wild-type bacteriocin-producing strain Wild-type bacteriocin-producing strain
ATCC 19321 (from M. Faelen) NCPPB 802 (from M. Faelen) NCPPB 577-8 (from M. Faelen)
E. herbicola 651
Wild-type bacteriocin-producing strain
Laboratoire Botanique, University of Liege (from M. Faelen)
Bacteriophages Mu cts62 Mu cts62 S::Tn9 Mu cts62 7701
Thermoinducible Produces only G(-) particles Produces only G(+) particles Mu dl BX::Tn9 (lac Apr Cm')
30 13 Thompson and Howe (unpublished) (from A. Toussaint) 3
Apr Tcr Cmr
10
JLV 634, JLV 635, JLV 636 JLV 637, JLV 638, JLV 639, JLV640, JLV641, JLV 642 E. coli CAG 5050 BHB 2600
E. chrysanthemi B374 1277, 1456, 1500, 1521, 1871, 1884
This work This work
E. uredovora
Mu dX
Plasmid pBR325 a
HR- does not induce hypersensitivity on tobacco plant.
sensitive (as shown by the presence of a turbid zone) to the bacteriocins produced by strains 20D3 and 20D31. No inhibition was observed with undiluted bacteriocin solutions from E. chrysanthemi 1277, 1456, 1500, 1521, 1871, or 1884. Sensitivity to this set of bacteriocins was used to confirm the identity of strain 1430. Preparation of bacteriophage lysates. Mu bacteriophage lysates were prepared from lysogenic strains of E. coli by thermal induction as described by Toussaint and Schoonejans (57). Plate lysates were made in SM buffer (61), as described by Toussaint and Schoonejans (57). Suspensions were sterilized by addition of a few drops of chloroform and centrifuged to remove cell debris. The titer of Mu stocks was determined by spotting 10-pul drops of 1/10 serial dilutions onto overlays seeded with either E. chrysanthemi B374 or E. coli BHB 2600 and incubated at 35 and 42°C, respectively. Isolation of lysogenic strains of E. amylovora. Mu infection was performed as described by Baker et al. (3). The strain of E. amylovora to be infected was grown to 2 x 108 cells per ml in L broth supplemented with 2.5 mM CaCl2 and 5 mM MgSO4. A fresh phage lysate was added to obtain a multiplicity of infection of the helper phage of between 0.1 and 0.05. Phages were allowed to adsorb at room temperature for 30 min. The infected cells were then diluted 1:10 with L
broth and incubated with shaking at 27°C for 30 min to allow phenotypic expression. Selection of lysogenic clones of strain 1430, after infection with a Mu dX lysate, was based on the chloramphenicol resistance conferred by the Tn9 carried by the Mu dX prophage. The frequency of lysogenization was calculated as the number of bacteria that acquired the antibiotic resistance divided by the total number of bacteria infected. Identification of mutants affected in pathogenicity. The screening was done by individual inoculation of apple root calli. The calli were propagated as described by Paulin and Duron (43). Calli, 3 to 4 weeks old, were cut into quarters just before inoculation. The strains to be tested were transferred from a 24-h plate with a toothpick and applied to the freshly made cut. Calli were incubated in the dark at 17°C. After 3 days of incubation, when exudate produced by the wild-type strain could already be seen, the calli were examined daily for the presence of bacterial exudate under x6 magnification. When no exudate was detected after 13 days of incubation, the strain was retested at least two more times. Pathogenicity and virulence tests. Pathogenicity tests were done on young plants of Malus domestica cv. Golden Delicious, Pyrus communis cv. Passe Crassane, and Pyra-
934
VANNESTE ET AL.
cantha cv. Rogersiana. Inocula were prepared by suspending bacteria, grown for 12 to 16 h on L agar plates, in sterile water to a final concentration of about 5 x 108 CFU/ml. We inoculated plants by cutting through the main vein of a young, fully developed leaf with scissors previously dipped in the bacterial suspension. Routinely, three plants with two leaves inoculated per plant were used per strain. When virulence was quantified, a 10-,ul drop of the bacterial suspension was dispensed between two parallel cuts made on the main vein of a young, fully developed leaf. This test was done on 10 plants per strain and repeated at least three times. Sterile distilled water was used as a control. Plants were maintained in a growth chamber at 21 to 24°C under 10,000 lux with a 12-h light, 12-h dark cycle. Under these conditions the wild-type strain 1430 produced typical symptoms within 2 days, followed by necrosis of the infected tissues. When immature pear fruits were used for the pathogenicity test, we followed the procedure described by Steinberger and Beer (55). HR on tobacco. HR-inducing ability was tested on leaves of the tobacco cultivar Xanthi. Inocula were prepared in sterile distilled water as described for the pathogenicity test. Infiltration of the intercostal area of young and fully expanded leaves was done with 18-gauge sterile needles. Sterile water was used as a control. Routinely, three infiltrations were done per strain. The tobacco plants were kept in a growth chamber under the same conditions as described for the pathogenicity and virulence tests. Under these conditions the wild-type strain 1430 developed a typical HR within 12 h. Isolation of virulent revertants. Virulent revertants of the mutants of classes 1 and 2 were recovered from drops of exudate produced on immature pear fruits inoculated with high bacterial populations. Immature pear fruits of cultivar Bartlett kept in the dark at 0 to 4°C with moderate aeration were surface disinfected with 70% ethanol. Slices were made and placed in plastic boxes on sterile paper towels previously moistened with sterile water. L broth (20 ml) was inoculated with a single colony of the mutant to be analyzed. Overnight cultures were centrifuged, and the pellet was suspended in 10 ml of 10 mM MgSO4 solution. Then 100 to 200 ,ul of this bacterial suspension was applied per slice. As soon as the bacterial suspension was absorbed, the fruits were incubated in the dark at 27°C for 12 days. We inoculated 40 to 70 pear slices per mutant. As a control, 50 pear slices were each inoculated with 100 to 200 ixl of 10 mM MgSO4 solution. After 2 weeks of incubation no exudate was detected. When a drop of ooze was detected on a pear slice inoculated with a mutant, it was immediately streaked out onto L medium plates. Twenty colonies per isolate were randomly chosen to determine the sensitivity to chloramphenicol. One colony resistant to chloramphenicol and one sensitive to chloramphenicol were used for pathogenicity and HR tests. DNA methods. Total DNA from E. amylovora was isolated as described by Klotz and Zimm (35) and purified by ethidium bromide-cesium chloride centrifugation (39). Plasmid DNA was extracted by using the procedure of Kado and Liu (33). When used as probe, the plasmid DNA was isolated from overnight cultures by a cleared-lysate procedure (15) and further purified by ethidium bromide-cesium chloride centrifugation (39). Mu DNA isolated from a lysogenic strain of Proteus mirabilis was a generous gift from A. Toussaint and M. Faelen, Universite Libre de Bruxelles, Brussels, Belgium. Restriction endonucleases were purchased from Bethesda Research Laboratories, Inc., Gaithersburg, Md.,
J. BACTERIOL.
and used as specified by the supplier. Phage and plasmid DNA were labeled by using a nick translation reagent kit purchased from Amersham Corp., Arlington Heights, Ill., with 100 LCi of [a-32P]dCTP used per reaction. The conditions for DNA labeling were those specified by the supplier. DNA electrophoresis was done as described by Maniatis et al. (39). The membranes used to transfer the DNA were either Biodyne 3G (Pall Biodyne) or GeneScreen Plus (Du Pont, NEN Research Products, Boston, Mass.). When Biodyne 3G membranes were used, DNA transfer and hybridization were carried out as described by Silhavy et al. (50). When using GeneScreen Plus we followed the instructions given by the manufacturer. Autoradiography was carried out at -80°C with XAR-5 film (Eastman Kodak Co., Rochester, N.Y.) Cronex Lightning-Plus intensifying screens (E. I. du Pont de Nemours & Co., Wilmington, Del.). Capsule staining. Capsule and slime layers were revealed by the wet India ink staining procedure of Duguid as described by Doetsch (20), with the following modifications. The strains to be tested were grown for 24 h on a carbohydrate-rich medium (YP medium). A colony was suspended onto a glass slide in a drop of sterile water and thoroughly mixed with a drop of India ink (Pelikan). Outer membrane preparations and analysis. Triton-insoluble walls (outer membranes) were prepared and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described by Expert and Toussaint (23). The proteins used as standards to calculate the apparent molecular sizes were purchased from Bio-Rad Laboratories, Richmond, Calif. LPS extraction and analysis. LPS was extracted from E. amylovora by using the miniscale procedure designed by Schoonejans et al. (48). Samples were analyzed by SDSPAGE on 13.5% acrylamide gels and stained by the silverstaining method of Hitchcock and Brown (28). Detection of a siderophore activity. Siderophore activity was detected by the chemical assay devised by Schwyn and Neilands (49). When plated onto the blue medium supplemented with nicotinic acid, colonies of strain 1430 were surrounded by an orange halo, indicating the production and excretion of a siderophore. RESULTS Strain 1430 is sensitive to phage Mu. The Mu genome contains a region, called G, that can invert. Each orientation of this DNA fragment corresponds to the synthesis of different proteins involved in the host specificity of the viral particle. The two different kinds of particles are called Mu G(+) and Mu G(-) according to the orientation of the G fragment in the Mu DNA (58). When a Mu cts62 lysate, containing both Mu G(+) and Mu G(-) particles, was spotted onto an overlay seeded with E. amylovora 1430, zones of partial inhibition arose within 16 h of incubation at 35°C. Mu lysates grown by infection of strain 1430 were made from these inhibition zones. The titer of these lysates was 1,000 times higher on E. chrysanthemi B374, which is sensitive to Mu G(-) particles, than on E. coli BHB 2600, which is sensitive to Mu G(+) particles. Since BHB 2600 is restriction negative and B374 was shown not to restrict the growth of Mu phages originating from different host species (57), the difference in plaque production observed between B374 and BHB 2600 suggests that strain 1430 is sensitive to Mu G(-) particles. Further evidence was sought by using the Mu cts62 S::Tn9 derivative that can produce only Mu G(-) particles and the Mu cts62 7701 derivative that can
NONPATHOGENIC Mu-INDUCED MUTANTS OF E. AMYLOVORA
VOL. 172, 1990
TABLE 2. Frequency of reversion to wild-type phenotype of auxotrophic Mu dX lysogenic derivatives of strain 1430 and sensitivity of the revertants to chloramphenicol and to Mu Mutant strain
JLV JLV JLV JLV JLV
522 523 524 525 526
Frequency of Axtoh
reesnto
Auxtotrophy thrvrint rototrophyof paA
Adenine Threonine Arginine Methionine Tryptophan
S 8 8 8 4
x
10-9
x 10-10 x 10-10 x 10-9 x
10-8
No. of
Cm' strains/no, tested
20/20 8/8 6/6 12/12 20/20
No. of Mu' strains/no. tested
5/5 2/2 2/2 3/3 NDa
a ND, Not determined.
produce only Mu G(+) particles. Although no growth on strain 1430 of either of these derivatives was detected, chloramphenicol-resistant derivatives of strain 1430 were recovered at a frequency of 7 x 10-6 when Mu cts62 S: :Tn9 lysates were used, whereas no lysogenic clones could be isolated when using Mu cts62 7701 lysates (frequency,
