Enhanced colonization and pathogenicity of ... - Wiley Online Library

Plant Pathology (2010) 59, 252–261

Doi: 10.1111/j.1365-3059.2009.02182.x

Enhanced colonization and pathogenicity of Erwinia amylovora strains transformed with the near-ubiquitous pEA29 plasmid on pear and apple M. Mohammadi Department of Plant Pathology, College of Agriculture and Natural Resources, University of Tehran, Karaj,31587-11167, Iran

Three plasmid-free strains of Erwinia amylovora, the causal agent of fire blight disease of pome trees, one from Iran, one from Egypt and one from Spain, were transformed with the near-ubiquitous nonconjugative pEA29 plasmid from a wildtype strain and characterized. The plasmid-deficient strains were levan- and slime-positive, motile, chemotaxis-positive, induced HR on Nicotiana tabacum var. xanthi but produced several-fold less amylovoran and were weakly pathogenic on pear slices and apple seedlings compared to plasmid-bearing wild-type strains. When inoculated onto wounded young apple (cv. Royal Gala) leaves, the plasmid-free strains labelled with green fluorescent protein gene (gfp) were mainly restricted to the inoculation site at the leaf tips, in contrast to the plasmid-carrying wild-type strains that moved into the midrib xylem vessel and colonized the adjacent parenchyma cells. Upon introduction of the transposon-labelled pEA29 plasmid, amylovoran production, degree of oozing and tissue necrosis on pear slices were significantly elevated in all three strains, whilst the levels of levan and levansucrase declined. Only the strains from Iran and Egypt gained the ability to invade and colonize the young apple leaves following the introduction of pEA29. It is concluded that acquisition of the nonconjugative near-ubiquitous plasmid may not necessarily confer increasing pathogenicity in all bacterial strains. Keywords: fire blight, pome fruit trees, pathogenicity

Introduction Erwinia amylovora causes fire blight on apple (Malus domestica), pear (Pyrus communis) and other rosaceous plants. The bacterial pathogen invades its hosts and causes tissue necrosis by entering the full-bloom blossoms via nectarthodes during spring when temperature and humidity are conducive to infection. The fire blight agent subsequently migrates downward through the xylem elements into the peduncle, shoot, leaves, branches, trunk and even root, causing water-soaking, oozing, wilting, necrosis and blight of the infected tissues. Severe symptoms often lead to the eventual death of the diseased tree (Bonn & van der Zwet, 2000). Factors contributing to the pathogenicity of E. amylovora on its host species include the production of amylovoran, an extracellular acidic heteropolysaccharide, and levan, a neutral polyfructan whose synthesis is catalysed by levansucrase, and the presence of functional hrp gene cluster. Mutations in the ams operon encoding amylovoE-mail: [email protected] Current address: Department of Plant Pathology and Microbiology, 2465 Boyce Hall, 3401 Watkins Drive, University of California-Riverside, Riverside, CA 92521, USA Published online 10 November 2009

252

ran or the hrp gene cluster result in strains failing to induce fire blight symptoms in host plants. Levan-deficient mutants were less aggressive (described as less virulent; Oh & Beer, 2005). Almost all wild-type strains of E. amylovora investigated to date do possess a nontransmissable indigenous plasmid known as pEA29 (Falkenstein et al., 1988; Laurent et al., 1989), except in one case in which the 29-kb ubiquitous plasmid was shown to be lacking but replaced by another native plasmid in a Spanish strain (Llop et al., 2006). Some strains of E. amylovora that are streptomycin-resistant carry the conjugative plasmid pEA34 containing the tandem strA-strB genes on Tn5393 (McManus & Jones, 1994). Because of its extreme stability and ubiquity in all pathogenic strains, pEA29 has been the focus of attention and thought to influence host biochemical and molecular properties as well as pathogenicity. The complete nucleotide sequence of the pEA29 plasmid in strain Ea88 revealed the existence of up to 21 open reading frames whose products show resemblance to the thiamin biosynthetic genes thiO, thiG and thiF and several other genes that encode for choline transport, methionine sulfoxide reductase, chemotaxis, a LysR-type transcriptional regulator and partitioning (McGhee & Jones, 2000). Plasmid-cured strains of E. amylovora showed attenuated pathogenicity on pear slices and pear seedlings, as well as modified exopolysaccharide (EPS) ª 2009 The Author Journal compilation ª 2009 BSPP

Transformation of Erwinia amylovora

production by forming translucent slime on minimal agar medium containing sucrose but without thiamin (Falkenstein et al., 1989; Laurent et al., 1989; McGhee & Jones, 2000). Molecular detection and identification of E. amylovora strains has been based on the amplification of a 1Æ0-kb PstI fragment of the common pEA29 plasmid (Bereswill et al., 1992; McManus & Jones, 1994; Llop et al., 2000), as well as a 1Æ6-kb genomic fragment amplified from the amsB gene involved in amylovoran synthesis (Bereswill et al., 1995) using specific primer pairs. The latter technique is also useful for identifying strains lacking the native plasmid. The main objective in this study was to investigate whether the near-ubiquitous plasmid plays any significant role in pathogenicity of E. amylovora. Hence, naturally occurring plasmid-deficient strains were transformed with the near-ubiquitous plasmid from a wildtype strain and their in planta invasion and colonization and other phenotypes were characterized.

Materials and methods Bacterial strains and growth media Bacterial strains used in this study are listed in Table 1. All E. amylovora strains were routinely cultured and maintained on Standard I nutrient agar (StdI) (37 g L)1, pH 7Æ5) medium (Merck). Bacterial colony morphology and growth characteristics were evaluated on minimal agar media containing asparagine and copper sulphate (MM1Cu, MM2Cu), MM2C and Luria-Bertani (LB)

253

sucrose plates prepared according to Bereswill et al. (1998). Minimal media (MM) supplemented with 2% (w ⁄ v) sucrose with or without thiamin were prepared according to Falkenstein et al. (1989).

PCR assays A bacterial colony was scraped off a StdI culture plate with a sterile plastic tip, resuspended in 0Æ5 mL sterile filtered Millipore water containing 1% (v ⁄ v) Tween20 in an Eppendorf tube and incubated at 65C for 30 min. The PCR mixture in a final volume of 50 lL consisted of 19 lL sterile filtered Millipore water, 25 lL 2· buffer [134 mM Tris HCl, pH 8Æ8; 3 mM MgCl2; 20 mM mercaptoethanol; 320 lg BSA mL)1; 10% (v ⁄ v) DMSO; 0Æ4 mM dNTP (Roche Diagnostics); 32 mM NH4(SO4)2], forward and reverse primers at 25 pmol lL)1 each, 2 units TaqDNA polymerase (PeqLab) lL)1 and 3 lL lysed bacterial cells. The primer sequences and PCR programmes for the plasmid pEA29 and 1Æ6-kb amsB chromosomal fragment were described previously (Bereswill et al., 1992, 1995). The PCR products were resolved in 1% (w ⁄ v) LE agarose gel in TAE buffer at 100 volts for 40 min, stained with ethidium bromide and photographed under UV light (302 nm). A GeneRuler 1-kb DNA ladder (MBI Fermentas) was used as a molecular size marker in gel electrophoresis analysis.

Electroporation of pEA29Tc and gfp plasmids Plasmid pEA29Tc labelled by random transposition with plasmid pUCD623 (Shaw et al., 1988) delivering Tn4431

Table 1 Bacterial strains and plasmids used in this study Strain

Origin, year

Host

Source or Reference

Ea1 ⁄ 79 Ea1 ⁄ 79sm EaX1 ⁄ 79sm Ea-Irn2 Ea-Irn37 Ea-Irn37+pEA29Tc Ea-A4 Ea-A4+pEA29Tc IVIA-1614-2e IVIA-1614-2e+pEA29Tc EaMR1 EaIL6 E. tasmaniensis (Et ⁄ 99) E. billingeae (Eb661sm) E. pyrifoliae (Ep1 ⁄ 96) E. pyrifoliae (Ep1 ⁄ 96 ⁄ C1) E. amylovora (MR1 ⁄ C1) E. amylovora (IL6 ⁄ C1) Pantoea stewartii (DC283) Escherichia coli (DH5a) Plasmids pfdC1Z’-gfp pEA29Tc

Germany, 1979 SmR strain of Ea1 ⁄ 79 Ea1 ⁄ 79sm cured of pEA29 Shahriar, Iran, 2002 Tabriz, Iran, 2001 Behera, Egypt, 1982 Segovia, Spain, 1996 Michigan, USA Illinois, USA Epiphyte, Tasmania, 1999 Epiphyte, England South Korea, 1996 Ep1 ⁄ 96 strain with gfp MR1 strain with gfp IL6 strain with gfp USA -

Cotoneaster sp. Apple Pear Pear cv. Le Conte Pyracantha sp. Raspberry Raspberry Apple, pear Pyrus pyrifolia (Nashi pear) Zea mays -

Falkenstein et al., 1988 ‘‘ ‘‘ ‘‘ ‘‘ This study This study This study K. Geider This study Llop et al., 2006 This study Jock & Geider, 2004 C. Bazzi via K. Geider Geider et al., 2006 Mergaert et al., 1999 Rhim et al., 1999 This study This study This study Coplin et al., 2002 Gibco BRL


gfp in pfdC1Z’, fd ori, KmR pEA29::Tn4431-30

Bogs et al., 1998 Bellemann et al., 1990

254

M. Mohammadi

(a Tn3 type, Tcr; Bellemann et al., 1990) was used to transform E. amylovora strains lacking the pEA29 plasmid via electroporation (Metzger et al., 1994). Transformants were selected on tetracycline plates and screened for the presence of pEA29 by PCR using A ⁄ B primers and plasmid profile. Erwinia amylovora strains Ea1 ⁄ 79Sm, EaX1 ⁄ 79Sm, Ea-Irn2, Ea-A4, Ea-A4+pEA29Tc, Ea-Irn37, Ea-Irn37+pEA29Tc, IVIA-1614-2e and IVIA1614-2e+pEA29Tc, as well as MR1 and IL6 from Rubus and E. pyrifoliae (Ep1 ⁄ 96) from Japanese pear, were transformed with pfdC1Z’ (KmR) plasmid DNA containing the green fluorescent protein gene (gfp; Bogs et al., 1998) by electroporation as follows. All strains were grown in StdI medium at 28C overnight then washed three times in sterile deionized water; the pellet was finally resuspended in 100 lL sterile water and mixed with 2 lg plasmid DNA on ice. The mixture was transferred into an electroporation cuvette and electroporated at 800 W, 25 lF and 2Æ5 Kv for 4 s using a Bio-Rad Gene Pulser. Following electroporation, 400 lL LB medium were added to each cuvette and the mixture then transferred into a fresh tube and incubated for 2 h on a shaker at 28C. One hundred lL of 10-fold dilutions of cell culture were plated out on StdI plates containing 20 lg kanamycin mL)1. To ensure the insertion and stability of the GFP marker, transformants were subcultured several times on StdI plates supplemented with 20 lg kanamycin mL)1.

Pathogenicity tests Immature pears (cv. Bartlett) were briefly washed in running tap water, air-dried, surface-sterilized in 70% ethanol and then cut into three slices, each 5 mm thick. Pear slices in triplicate were placed in a deep Petri dish (69 · 22 mm, neoLab GmbH) and stab-inoculated once with a sterile toothpick dipped into a fresh bacterial colony. Plates were incubated at 28C for 1 week and scored for ooze production and tissue necrosis. Apple seedlings (Malus rootstock) in triplicate were also used in pathogenicity tests. Young leaves were inoculated in shoot tip with a sterile fresh toothpick dipped into a cell suspension (109 cells per mL) using the above technique. Plants were incubated at 28C in a greenhouse and scored for fire blight symptoms 2 weeks later. Strains grown overnight in 1 mL StdI medium containing 20 lg kanamycin mL)1 were centrifuged and the pellet washed in sterile water several times and finally resuspended in 1 mL water. The optical density (OD600nm) was adjusted to 1Æ0. Apple plantlets (cv. Royal Gala) were prick-inoculated in shoot tips using a sterile toothpick dipped in the cell suspension. Bacterial invasion of plant tissue at the wound site and migration into xylem vessels and parenchyma cells were evaluated 5 days after inoculation at room temperature using an epifluorescence microscope (Axiovert S100 Zeiss) with the filter combination BP450-490 ⁄ FT510 ⁄ LP520 (excitation filter ⁄ dichroic-emission filter) and a · 10 magnifi-

cation lens. Bacterial spread was scored on a scale of 0–5 as the distance from the point of inoculation at the leaf tip, 0 representing no migration into leaf tissue and 5 being the invasion of the entire xylem vessel and the adjacent parenchyma cells. Hypersensitivity was tested by growing strains in 1 mL StdI medium on a shaker at 28C overnight, resuspending the centrifuged pellet in 1 mL sterile distilled water and, after adjusting the turbidity to 1Æ0 (OD600nm), infiltrating 1 · 108 cells per mL into young tobacco (N. tabacum cv. Samson) leaves using a 1-mL sterile syringe (BD Plastipak). Hypersensitive reaction symptoms were scored on triplicate leaf panels after 5 days of incubation at room temperature.

Amylovoran assay Amylovoran synthesis was measured by growing bacterial cells in 4 mL MM2C medium in small flasks on a rotary shaker at 28C for 48 h. Flasks were initially inoculated from overnight colonies grown on StdI plates. Subsequent subculturings were carried out by taking 10 lL from a previous culture and transferring it into fresh medium. Cell cultures in 1Æ5-mL tubes were centrifuged at 15 871 g for 5 min. Culture supernatant (50–500 lL) was mixed with 50 lL of 5% (w ⁄ v) cetylpyridinium chloride (CPC) (Sigma) in a total volume of 500 lL MM2C buffer in a plastic cuvette, then the mixture incubated for 10 min at room temperature. Absorbance was read at 600 nm against an MM2C blank. The amount of amylovoran produced was extrapolated from a standard calibration curve prepared using a commercial EPS. Results were expressed as lg amylovoran ⁄ OD cell density. In case the amount of amylovoran in the culture supernatant was too high, appropriate dilutions were prepared in MM2C medium and mixed with 5% CPC for each strain.

Levansucrase assay Levansucrase secreted into the growth medium was assayed spectrophotometrically following the method of Bereswill & Geider (1997) with slight modification. Cells were grown in 1 mL LB medium at 27C for 48 h and centrifuged at 15 871 g for 15 min. A 250-lL volume of supernatant was mixed with an equal volume of 2· levansucrase buffer (50 mM sodium phosphate (pH 7Æ0), 2 M sucrose and 0Æ05% sodium azide) in a plastic cuvette and incubated for 24 h at 37C. Turbidity caused by levan accumulation was measured at 600 nm.

Siderophore agar plate assay Fresh bacterial colonies on StdI medium were picked by a sterile dissecting needle and stabbed into a chrome azurol (CAS) agar plate prepared according to Schwyn & Neilands (1987). Plates were incubated at 27C for 48 h and the diameter of orange halo (siderophore excretion) around the colonies on blue agar was measured. Plant Pathology (2010) 59, 252–261


6 4 2 0

(b) 100 75 50 25

Chemotaxis experiment

0 (c) 75 Oozing (%)

Bacterial strains were stab-inoculated deep into minimal medium (4 g agar L)1) without sucrose (Falkenstein et al., 1989) supplemented with 100 mM L-aspartic acid, L-serine and succinic acid or 1 mM DL-malic acid, pH 7Æ2 (McGhee & Jones, 2000). Bacterial swarming was evaluated by measuring colony diameter (mm) after 2, 4 and 6 days of incubation at 28C.

50 25

Data analyses

Figure 1 Factors contributing to pathogenicity in strains of Erwinia amylovora with (bold) and without the pEA29 plasmid. (a) Levan production on LB agar supplemented with 5% (w ⁄ v) sucrose; mean values and standard errors of eight replicates from two independent experiments. (b) Amylovoran synthesis; mean values and standard errors of 10 replicates from five independent experiments. (c) Percentage area of immature pear fruit slices covered with bacterial ooze; mean values and standard errors of 10–15 replicates from three or four independent experiments. (d) Tissue necrosis in pear slices; mean values and standard errors of 10–15 replicates from three or four independent experiments. (e) Sensitivity to 300 mM H2O2 in disc assay. Mean values and standard errors of 3–13 independent assays.


Tissue necrosis (%)

0 (e) Radius of inhibition zone (mm)

Two plasmid-free strains of E. amylovora, Ea-A4 and EaIrn37, isolated from pear trees with fire blight symptoms in Egypt and Tabriz, Iran, respectively, were identified

25

8 6 4 2 0 79

Growth characteristics

50

1/

Results

Ea

The data from each experiment were analysed using Student’s t-test. Details of experimental replication and repetition are given in Figs 1 and 2.

0 (d) 75

s X m 1/ 79 s Ea m -I rn Ea Ea 2 -A -A 4+ 4 pE A Ea 29 Ea T -I rn -Ir c 37 n3 +p 7 IV E IV A IA 2 I A 9T -1 -1 61 61 c 442e 2e +p EA 29 Tc

Bacterial colonies were stab-inoculated deep into the centre of MOM (motility medium) agar plates (3 g agar, 7 g NaCl, 1Æ15 g K2HPO4, 0Æ2 g KH2PO4, 0Æ2 g KCl, 0Æ1 g yeast and 0Æ05 g sucrose L)1). Swarming motility was monitored and colony diameter measured in triplicate after 24, 48, 72 and 89 h of incubation at 28C.

8

Ea

Swarming motility test

(a) 10

Colony diameter (mm)

Bacterial strains were cultured in 5 mL Ceria broth to an OD600nm of 0Æ5 at 27C following the protocol of McGhee & Jones (2000) with slight modification. Aliquots of 500 lL were mixed with 5 mL MM+FeSO4 soft agar at 50C and poured over the surface of MM+FeSO4 plates. After 1 h of incubation at 20C, five sterile filter discs (10 mm in diameter) were placed in the centre of each plate and 100 lL of 300 mM hydrogen peroxide added to the filter discs. Plates were incubated at 27C overnight and evaluated for zones of growth inhibition.

and characterized based on phenotypical and molecular properties and compared to those of plasmid-containing reference strains, Ea1 ⁄ 79Sm from Germany and Ea-Irn2 from Iran (Table 1). A pEA29-deficient strain, IVIA-

µg EPS/OD

Oxidative stress disc inhibition assay

255

Strain

M. Mohammadi

agarose gel and specific primers used to amplify DNA fragments from the pEA29 plasmid and chromosome in PCR assays. The pEA29 plasmid was present in the wildtype strains Ea1 ⁄ 79 and Ea-Irn2, but absent from strains Ea-A4, Ea-Irn37 and IVIA-1614-2e (Table 2). In addition, unlike strains Ea1 ⁄ 79Sm and Ea-Irn2, it was not possible to amplify the 1Æ0-kb PstI fragment of the pEA29 plasmid from strains Ea-A4, Ea-Irn37 and IVIA-1614-2e, confirming the lack of the pEA29 plasmid in these strains (Table 2). Further identification of E. amylovora strains devoid of plasmid pEA29 was carried out using specific primers to amplify a chromosomal fragment from the amsB region. Similar to the wild-type strains, pEA29deficient strains amplified the corresponding 1Æ6-kb amsB fragment (Table 2).

0·6 0·4 0·2

Tc

2e

29 EA

61

29 IV

IA

-1

Ea -

61

4-

IV

2e

IA

+p

-1

EA +p

37 Ir n

A Ea -

4-

Tc

7 n3

Tc Ea

-Ir

29 pE A

4+

Ea

-A

4

0 Ea 1/ Ea 79S m X 1/ 79 Sm

OD600nm/OD cell density

256

Strain Figure 2 Levansucrase activity in Erwinia amylovora strains with (bold) or without pEA29 plasmid transformation. Values are the means and standard errors of six replicates from three separate experiments per OD600nm cell density.

Exopolysaccharide production

1614-2e, isolated from Pyracantha sp. in Segovia, Spain, was also included for comparison. All five strains were able to form yellow colonies on plates of MM2Cu medium, which is specific for morphological and cultural identification of E. amylovora (Table 2). All strains had mucoid colony appearance on MM2Cu, except strain Ea-A4, but none were capable of growing on MM1Cu medium, typical of E. amylovora. Strains varied with respect to slime production on MM2C plates, levan synthesis on LB sucrose plates and levansucrase secretion in LB medium. All strains produced slimes, with the exception of Ea-A4.

Molecular confirmation of pEA29 plasmid In order to confirm the presence or absence of the nearubiquitous pEA29 plasmid in E. amylovora strains, plasmid DNA was extracted and electrophoresed in an

The role of the near-ubiquitous plasmid in pathogenicity was studied following the introduction of transposonlabelled pEA29 (pEA29Tc) derived from Ea1 ⁄ 79Sm into strains lacking pEA29. All three strains transformed with pEA29 formed yellow mucoid colonies on MM2Cu with fluidal morphology and with enhanced production of slime on MM2C agar plates (Table 2). However, levan synthesis was reduced significantly, by 23, 10 and 20% for Ea-A4+pEA29Tc (P < 0Æ005), Ea-Irn37+pEA29Tc (P < 0Æ05) and IVIA-1614-2e+pEA29Tc (P < 0Æ01), respectively (Fig. 1a). Likewise, there was a substantial drop in levansucrase activity following acquisition of the plasmid. Levansucrase secretion diminished significantly (P < 0Æ005) in the transformed strains of Ea-A4, Ea-Irn37 and IVIA-1614-2e by 36, 64 and 71%, respectively (Fig. 2). Levansucrase activity in Ea1 ⁄ 79Sm was low and curing of the pEA29 plasmid led to an increase in levansucrase activity by 57%. Both levan and amylovoran are exopolysaccharides (EPS) that contribute to in planta pathogenicity of E. amylovora (Oh & Beer, 2005). Amylovoran synthesis

Table 2 Growth characteristics, molecular properties, degree of disease severity on apple seedlings and HR in tobacco of Erwinia amylovora strains, some transformed with pEA29Tc

Strain

MM2Cua

MM1Cu

MM2Cb

pEA29 plasmid

1Æ0-kb PstI fragment

1Æ6-kb amsB fragment

Disease severity on apple seedlingsc

HRd

Ea1 ⁄ 79sm EaX1 ⁄ 79sm Ea-Irn2 Ea-A4 Ea-A4+pEA29Tc Ea-Irn37 Ea-Irn37+pEA29Tc IVIA-1614-2e IVIA-1614-2e+pEA29Tc

+ + + + + + + + +

) ) ) ) ) ) ) ) )

+ + + ) ) + ++ + ++

+ ) + ) + ) + ) +

+ ) + ) + ) + ) +

+ + + + + + + + +

+++ ++ ++ ++ +++ ++ +++ + ++

+ + + + + + + + +

(M) (M) (M) (M) (M) (M) (M)

a

+, Yellow colony formation on MM2 cupper sulfate plate incubated at 28C for 5 days. M, mucoid to fluidal at rim of colony. Slime production on MM2C plate: +, low; ++, moderate, both with raised colonies; ), flat colony with a crater. c Degree of disease severity: +, weak; ++, moderate; +++, high. Results are the average of three apple seedlings after separate inoculation with each individual bacterial strain. The length of blighted shoot was measured as the degree of disease severity. d Hypersensitive reaction. b



increased dramatically in two strains after electroporation with pEA29 – by 79% in Ea-Irn37+pEA29Tc (P < 0Æ05) and by 138% (P < 0Æ005) in the Spanish strain, IVIA-1614-2e+pEA29Tc. In contrast, amylovoran level remained constant in Ea-A4 despite the introduction of pEA29 (Fig. 1b). Curing of pEA29 from Ea1 ⁄ 79Sm led to a 27% reduction in amylovoran production (P < 0Æ0005).

Symptom expression All five E. amylovora strains induced the hypersensitive reaction (HR) on tobacco leaves and caused typical fire blight symptoms such as shepherd’s crook, oozing and shoot, stem and leaf blight on apple seedlings (Table 2). However, disease severity differed among the strains, with Ea1 ⁄ 79Sm being the most pathogenic, followed by Ea-Irn2. The pEA29-deficient strains Ea-A4 and EaIrn37 caused intermediate disease, whilst IVIA-1614-2e was the least pathogenic. Following the introduction of pEA29Tc into strains lacking pEA29, they showed a higher degree of severity on apple seedlings, comparable to that of the wild-type strains (Table 2). Inoculation of immature pear slices with strains with and without pEA29 caused ooze production, water soaking and tissue necrosis. The degree of pathogenicity was, however, variable between the two groups of strains. Plasmid-bearing strains such as Ea1 ⁄ 79Sm and Ea-Irn2 were more aggressive and formed more ooze on pear slices than less aggressive strains deficient in pEA29 (Fig. 1c). Furthermore, there was an increase in the degree of oozing as well as the extent of tissue necrosis on

257

pear slices with all three pEA29-transformed strains compared with the corresponding wild-type parental strains (Fig. 1d). Bacterial oozing significantly increased by 63, 76 and 66% in Ea-A4 (P < 0Æ0025), Ea-Irn37 (P < 0Æ01) and IVIA-1614-2e (P < 0Æ01), respectively, following plasmid acquisition, and tissue necrosis by 74 (P < 0Æ15), 95 (P < 0Æ10) and 335% (P < 0Æ005), respectively, compared with the corresponding parental strains lacking the pEA29 plasmid. The wild-type Ea1 ⁄ 79Sm was the most aggressive strain on pear slices, inducing oozing and necrosis, followed by the less aggressive EaIrn2 strain. Curing of the pEA29 plasmid from Ea1 ⁄ 79Sm resulted in a significant reduction (P < 0Æ0005) in tissue necrosis.

In planta migration Spread of gfp-labelled E. amylovora strains in young apple leaves was evaluated (Table 3). Compared with Ea1 ⁄ 79Sm and Ea-Irn2, the pEA29-deficient strains Ea-A4, Ea-Irn37 and IVIA-1614-2e showed limited invasion into the midrib vein and parenchyma and in most cases fluorescing cells were restricted to the inoculation site. Migration indices of Ea-A4, Ea-Irn37 and IVIA1614-2e into the xylem vessels of young apple leaves were 3Æ3-, 3Æ8- and 10Æ4-fold less than those of the plasmid-carrying Ea1 ⁄ 79Sm and Ea-Irn2 strains, respectively. Similarly, migration indices of plasmid-free strains Ea-A4, Ea-Irn37 and IVIA-1614-2e from the xylem vessels into the surrounding parenchyma cells were 3Æ5-, 4Æ7- and 20Æ6- fold slower, respectively, than those of strains Ea1 ⁄ 79Sm and Ea-Irn2.

Table 3 Migration of plasmid-free and plasmid-transformed Erwinia amylovora strains labelled with GFP ⁄ C1 gene into young apple (cv. Royal Gala or Elstar) or pear (cv. Passe Crassan) leaves Strain

Host ⁄ cultivar

No. leaf samples

No. infected

Infected (%)

V(SE)

P(SE)a

Ea1 ⁄ 79Sm ⁄ C1 EaX1 ⁄ 79Sm ⁄ C1 Ea-Irn2 ⁄ C1 Ea-Irn2 ⁄ C1 Ea-A4 ⁄ C1 Ea-A4 ⁄ C1+pEA29Tc Ea-Irn37 ⁄ C1 Ea-Irn37 ⁄ C1+pEA29Tc IVIA-1614-2e ⁄ C1 IVIA-1614-2e ⁄ C1+pEA29Tc Ep1 ⁄ 96 ⁄ C1 Ep1 ⁄ 96 ⁄ C1 Ea-MR1 ⁄ C1 Ea-MR1 ⁄ C1 IL6 ⁄ C1 IL6 ⁄ C1

Royal Royal Royal Elstar Royal Royal Royal Royal Royal Royal Pear Royal Royal Pear Royal Pear

116 84 125 116 99 86 116 227 120 136 42 72 122 76 73 74

105 27 81 53 75 39 95 210 18 0 33 53 45 21 20 2

90Æ5 32 65 46 75Æ7 45 82 92Æ5 15 0 78Æ6 73Æ6 36Æ9 27Æ6 27Æ4 2Æ7

1Æ25 0Æ15 1Æ27 1Æ44 0Æ38 0Æ99 0Æ33 1Æ77 0Æ12 0 2Æ34 0Æ91 0Æ06 0Æ27 0Æ03 0Æ01

1Æ03 (0Æ14) 0Æ70 (0Æ01) 1Æ0 (0Æ15) 0Æ98 (0Æ21) 0Æ29 (0Æ09) 0Æ97 (0Æ3) 0Æ22 (0Æ06) 1Æ62 (0Æ13) 0Æ05 (0) 0 2Æ1 (0Æ37) 0Æ27 (0Æ09) 0Æ03 (0Æ0) 0Æ06 (0Æ05) 0Æ02 (0Æ0) 0Æ01 (0)

Gala Gala Gala Gala Gala Gala Gala Gala Gala Gala Gala Gala

(0Æ15) (0Æ04) (0Æ18) (0Æ25) (0Æ10) (0Æ28) (0Æ08) (0Æ13) (0Æ01) (0Æ36) (0Æ15) (0Æ024) (0Æ24) (0) (0)

a Inoculated plants (1–2 months old) were incubated at room temperature for 5 days and bacterial migration into the midrib xylem vessel (V) and adjacent cortical parenchyma cells (P) scored on a scale of 0–5, with 0 = no migration, 1 = 20%, 2 = 40%, 3 = 60%, 4 = 80%, 5 = 100% colonization. Numbers in parentheses are standard errors (±SE). Student’s t-test revealed significant differences in bacterial invasion of young apple leaves between plasmid-free and plasmid-transformed strains: Ea1 ⁄ 79Sm ⁄ C1 vs. EaX1 ⁄ 79Sm ⁄ C1 (V, P < 0Æ0005; P, P < 0Æ15); Ea-A4 ⁄ C1 vs. Ea-A4 ⁄ C1+pEA29Tc (V, P < 0Æ01; P, P < 0Æ005); Ea-Irn37 ⁄ C1 vs ⁄ Ea-Irn37 ⁄ C1+pEA29Tc (V, P < 0Æ0005; P, P < 0Æ0005).


258

M. Mohammadi

Strain

L-Aspartic

Ea1 ⁄ 79sm EaX1 ⁄ 79sm Ea-A4 Ea-A4+pEA29Tc Ea-Irn37 Ea-Irn37+pEA29Tc Ea-IVIA1614-2e Ea-IVIA1614-2e+pEA29Tc

13Æ8 (4Æ1) 10Æ2 (1Æ6) 15Æ2 (2Æ6) 9Æ2 (0Æ8) 9Æ9 (1Æ3) 4Æ8 (0Æ4) 13Æ5 (1Æ0) 7Æ2 (0Æ7)

acid

DL-Malic

acid

14Æ5 (1Æ0) 9Æ3 (1Æ1) 21Æ2 (0Æ2) 28Æ0 (1Æ3) 17Æ2 (2Æ2) 14Æ2 (1Æ5) 18Æ0 (0Æ8) 18Æ0 (0Æ6)

L-Serine

Succinic acid

5Æ4 5Æ0 6Æ5 6Æ7 5Æ5 8Æ0 7Æ0 8Æ3

7Æ4 (1Æ8) 9Æ0 (0) 12Æ2 (1Æ6) 12Æ3 (2Æ9) 6Æ5 (2Æ0) 6Æ5 (0Æ8) 8Æ0 (1Æ0) 7Æ0 (0Æ6)

(0Æ6) (0Æ3) (0Æ4) (0Æ3) (1Æ0) (1Æ5) (0Æ8) (0Æ9)

Table 4 Colony swarming (diameter in mm) of Erwinia amylovora strains on various organic compounds

a Plates were scored after 6 days of incubation at 27C. Each value represents the mean of three independent experiments. Numbers in parentheses are standard errors (±SE).

The movement of the wild-type strain Ea1 ⁄ 79Sm into the midrib of young apple leaves (cv. Royal Gala) was more than eightfold greater than when it was cured of its pEA29 plasmid. A second wild-type strain used as a positive control with pEA29 was Ea-Irn2, which invaded apple shoots (cvs Royal Gala and Elstar) to a similar extent as Ea1 ⁄ 79Sm. Migration of E. pyrifolia strain Ep ⁄ 1 was examined on both apple and pear leaves. This species, which infects Japanese pear, was highly aggressive on pear (cv. Passe Crassan), invading both the midrib and the adjacent parenchyma cells. Its invasion of apple leaves (cv. Royal Gala) was of the same magnitude as that of the wild-type Ea1 ⁄ 79Sm and Ea-Irn2 strains. MR1, which is exclusively pathogenic on Rubus (raspberry), was also tested on both apple cv. Royal Gala and pear cv. Passe Crassan. The Rubus strain was slightly pathogenic on pear leaves, but not apple leaves. The second negative-control pathogen on Rubus was IL6, which was unable to move into the young apple leaves and thus remained restricted to the wounded inoculation site. Introduction of plasmid pEA29Tc enabled both Ea-A4 and Ea-Irn37 to become significantly more intrusive and better colonizers of the apple midribs and parenchyma cells than IVIA-1614-2e+pEA20Tc, which was primarily a poor colonizer with or without plasmid transformation, as it remained at the inoculation site. Transformation with pEA29Tc allowed strains Ea-A4 and Ea-Irn37 to colonize apple midribs more than 2Æ5 and 5 times, respectively, and parenchyma cells more than 3 and 7 times, respectively, greater than in the absence of the plasmid.

Siderophore production Siderophore secretion, a potential factor for pathogenicity, was explored on CAS medium for wild-type and plasmid-cured as well as plasmid-free and plasmidtransformed strains. No substantial differences were observed in siderophore production with strains before and after introduction of the near-ubiquitous plasmid (data not shown). This was also true with Ea1 ⁄ 79Sm and its plasmid-cured derivative. Erwinia pyrifoliae Ep1 ⁄ 96, Erwinia billingeae Eb661sm and Escherichia coli DH5a did not secret any siderophore. In contrast, Erwinia tas-

maniensis Et ⁄ 99 and Pantoea stewartii DC283 were siderophore producers.

Oxidative stress Oxidative stress has been associated with the near-ubiquitous pEA29 plasmid in E. amylovora. Tolerance to hydrogen peroxide declined in Ea-A4, Ea-Irn37 and IVIA-1614-2e strains by 23, 27 and 219%, respectively, following the introduction of pEA29 (Fig. 1e), whilst curing Ea1 ⁄ 79Sm of plasmid pEA29 led to an 11% increase in tolerance to hydrogen peroxide. Ea1 ⁄ 79Sm was more tolerant to oxidative stress than Ea-Irn2. Only IVIA1614-2e+pEA29 became significantly more sensitive (P < 0Æ005) to hydrogen peroxide than the other transformed strains.

Swarming motility On MOM agar plate, strains devoid of the near-ubiquitous plasmid were as motile as plasmid-containing wildtype strains. Strains transformed with pEA29, however, were slightly less motile than the corresponding isogenic parental strains. There were no significant differences in swarming motility between strains before and after transformation with pEA29 (data not shown).

Chemotaxis Introduction of plasmid pEA29 into all three strains deficient in it led to a significant reduction in colony swarming on L-aspartic acid compared to the parental strains (Table 4). Ea-Irn37+pEA29 (P < 0Æ02) showed 50% loss of colony swarming on L-aspartic acid compared with 40% in Ea-A4+pEA29 (P < 0Æ05) and 47% in IVIA1614-2e+pEA29 (P < 0Æ0005). In Ea1 ⁄ 79Sm, plasmid curing slightly reduced colony swarming on L-aspartic acid. On DL-malic acid, there was a significant reduction in colony swarming in plasmid-cured Ea1 ⁄ 79sm (P < 0Æ02). In Ea-A4+pEA29Tc, colony swarming on DL-malic acid increased by 32% (P < 0Æ01) over that of the isogenic parental strain, whereas there were no significant changes in Ea-Irn37+pEA29 and IVIA-16142e+pEA29 strains. Plant Pathology (2010) 59, 252–261


No significant differences were observed in colony swarming between strains with or without pEA29 plasmid on L-serine and succinic acid.

Discussion The non-transmissible indigenous pEA29 plasmid in E. amylovora is known to be ubiquitous in almost all the strains characterized thus far, with the exception of strain IVIA-1614-2a from Segovia, Spain, which was reported to possess a 70-kb native plasmid instead (Llop et al., 2006). The 29-kb plasmid was entirely sequenced in strain Ea88 and several open reading frames were identified with nucleotide homology to genes for thiamine biosynthesis, choline transport, peptide methionine sulfoxide reductase, chemotaxis, a LysR-type transcriptional regulator and partitioning (McGhee & Jones, 2000). Two plasmid-deficient strains of E. amylovora, one from Iran and one from Egypt, isolated from fire-blightaffected pear trees, were characterized and compared to two pEA29-bearing strains, Ea1 ⁄ 79Sm and Ea-Irn2, isolated in Germany and Iran, respectively. Strain IVIA1614-2e was also included in this study for comparison. All three strains lacking pEA29 produced less bacterial ooze and induced less necrosis on immature pear slices than plasmid-bearing strains. Strains cured of pEA29 plasmid were previously shown to exhibit attenuated pathogenicity with a delay in symptom development on apple seedlings and pear slices, and slow migration and spread into the leaves (Falkenstein et al., 1989; Laurent et al., 1989; Geier & Geider, 1993; Aldridge et al., 1997). Introduction of pEA29Tc derived from Ea1 ⁄ 79Sm into Ea-Irn37 and IVIA-1614-2e resulted in a remarkable increase in the degree of amylovoran synthesis as well as virulence on apple seedlings and immature pear slices. Although the transformed Ea-A4 strain displayed an increase in aggressiveness on pear slices, its level of amylovoran production remained unchanged. Both levan and amylovoran are considered pathogenicity factors in E. amylovora, and mutation in the amylovoran operon resulted in the creation of strains that were nonpathogenic on pears (Metzger et al., 1994). According to Laurent et al. (1989), the pEA29 plasmid seems to play a quantitative role in pathogenicity. The exact mechanism underlying the contribution of pEA29 to the chromosomal expression of the ams operon responsible for exopolysaccharide synthesis in E. amylovora is not completely known. Maes et al. (2001) reported a correlation between in vitro amylovoran synthesis and the degree of pathogenicity exhibited by wild-type strains of E. amylovora on pear fruits. However, this was not the case with Ea-A4+pEA29Tc, which became more aggressive on pear slices without increasing its amylovoran production in vitro. This clearly shows that there might be other factors contributing to pathogenicity in addition to amylovoran. Slime production increased only in strains Ea-Irn37 and IVIA-1614-2e after the introduction of pEA29, Plant Pathology (2010) 59, 252–261

259

whilst levan production and levansucrase activity both declined in all three transformed strains. EPS synthesis in E. amylovora is regulated by the rcsA and rcsB genes, which are considered activators of amylovoran production (Bernhard et al., 1990). RcsC acts as a sensor, whereas RcsB is an activator protein in a two-component system enhanced by RcsA. Inactivation of rcsB resulted in amylovoran-deficient and nonpathogenic mutants in E. amylovora (Bereswill & Geider, 1997). Over-expression of the rcsA or rcsB gene significantly enhanced amylovoran synthesis and reduced levan-forming levansucrase activity, possibly altering cell metabolism, consistent with the findings in this study. It appears that the pEA29 plasmid can modulate synthesis of amylovoran, secretion of levansucrase and pathogenicity in the host plant depending on the environment and life style of its host. Whether pEA29 confers a fitness advantage to E. amylovora is not known. Although plasmids may carry genes beneficial to their bacterial hosts, they may also reduce fitness. The presence of the pEA29Tc plasmid appears to enhance the fitness advantage in some E. amylovora strains, but not all. Plasmid-free strains were substantially less aggressive in invading wounded young apple leaves than strains containing the near-ubiquitous plasmid; however, upon introduction of pEA29Tc, both Ea-A4 and Ea-Irn37 gained the ability to invade and colonize apple leaves to a significant level. Likewise, Ea1 ⁄ 79Sm cured of its native plasmid lost its capacity to invade young apple leaves by almost 10-fold. Plasmids are not always beneficial to their hosts. For instance, strain IVIA-1614-2e+pEA29Tc was unable to invade even the wounded edge of the leaf blade and as a result no fluorescent signal was observed. Whether pEA29 increases fitness to a host may depend on the genetic background of the recipient and possibly the presence of other plasmids. It is not clear why the transformed strain from Spain labelled with gfp was not as intrusive as the strains from Iran and Egypt bearing the pEA29 plasmid. One possible explanation could be the simultaneous presence of three plasmids in this strain following the transformation with both pEA29Tc from Ea1 ⁄ 79Sm and pfdC1Z’gfp and the existence of the 70-kb native plasmid. This may cause an adverse effect on bacterial mobility and colonization of young apple leaves, although the respective parental strain was also immobile. Alternatively, the inability of IVIA-1614-2e+pEA29Tc to invade young apple leaves might be related to its highly increased sensitivity towards hydrogen peroxide compared with strains Ea-A4+pEA29Tc and Ea-Irn37+pEA29Tc. Dellagi et al. (1998) and Venisse et al. (2003) demonstrated the necessity of desferrioxamine siderophore production by E. amylovora for tolerance to high levels of hydrogen peroxide in the apoplast. Thus, IVIA-1614-2e+pEA29Tc might be incapable of overcoming the combination of defences, stress and poor nutrition in apple leaf tissue as readily as other strains do. Despite the lack of intrusiveness, IVIA1614-2e+pEA29Tc was able to invade and colonize pear slices to a significant degree compared with the parental strain lacking pEA29. Such a discrepancy may lie in the

260

M. Mohammadi

physiological and biochemical differences that exist between pear fruit slices and young apple leaves. Availability of nutrients, iron and carbon sources in the intercellular spaces of apple leaves is perhaps limited compared to pear tissue slices rich in nutrients including iron. Erwinia amylovora must overcome these limitations in order to survive, multiply, colonize and cause symptoms. On other hand, Rico & Preston (2008) demonstrated that plant-pathogenic Pseudomonas syringae pathovars differed genetically in their abilities to utilize nutrients available in tomato leaf apoplast; apoplastic fluids upregulated the expression of the type-III protein secretion pathway during in vitro growth of P. syringae pv. tomato DC3000 compared with other pathovars. Therefore, it is possible that genetic background may also determine the ability of E. amylovora strains to utilize nutrients accessible in the apple leaf apoplast. McGhee & Sundin (2007) have reported that plasmidcured strains of E. amylovora complemented with the thiOGF operon from pEA29 restored pathogenicity in pear fruits and amylovoran synthesis on thiamine-deficient minimal medium. Furthermore, they showed that the thiOGF operon is necessary for enhanced expression of the ams operon during the infection of immature pear fruits. Zhao et al. (2005) demonstrated an upregulation of levansucrase and amylovoran biosyntheses during infection of pear fruits by E. amylovora. Similarly, the current study showed that pEA29Tc in the transformed strains contributed to enhanced amylovoran synthesis, and pathogenicity on pear slices and apple leaf colonization, although this was not consistent in all three strains. Plasmids have been shown to contribute to aspects of pathogenicity in other plant bacterial pathogens, e.g. Clavibacter michiganensis subsp. michiganensis (Meletzus et al., 1993), Xanthomonas oryzae pv. oryzae (Amuthan & Mahadevanm, 1993) and Agrobacterium tumefaciens (Nair et al., 2003). The methionine sulfoxide reductase gene (mrs), which is considered a virulence determinant in the plant pathogen Erwinia chrysanthemi (El Hassouni et al., 1999), was shown to exist on the pEA29 plasmid in E. amylovora as well (McGhee & Jones, 2000). The mrs gene encodes an enzyme which repairs proteins which have undergone oxidation in the pathogen as a result of oxidative burst in the host plant during a compatible interaction. The repair mechanism allows the pathogen to survive and colonize the host tissue. Although the pEA29 plasmid carries a gene encoding methionine sulfoxide reductase (McGhee & Jones, 2000), the acquisition of the near-ubiquitous plasmid rendered the E. amylovora strains less tolerant to oxidative stress, particularly in IVIA-1614-2e. Thus, it is likely that either the mrs gene is not fully expressed in the strains transformed with pEA29Tc, or it is partially suppressed, or it is only functional in planta. No substantial differences were observed in motility among the parental wildtype, pEA29-deficient and pEA29Tc-transformed strains. Furthermore, no differences in bacterial fitness were seen when cultures of pEA29Tc-transformed and

parental strains were compared under various laboratory conditions. Hence, the native plasmid in E. amylovora seems to be nonessential for successful growth and metabolism of its host. Incorporation of the pEA29 plasmid reduced colony swarming on L-aspartate in all three transformed strains, whereas there were no significant changes in chemotaxis occurring on L-serine and succinic acid. On DL-malic acid, only Ea-pEA29Tc showed a substantial increase in colony swarming. Thus, the introduction of the pEA29 plasmid did not enhance colony swarming in all three strains and transformed strains behaved uniformly towards the different amino acids, with the exception of DL-malic acid. This further suggests that acquisition of the nearubiquitous plasmid impacts the recipient strains differently and genetic background contributes to chemotaxis, motility and colony swarming to a significant extent. It would be interesting to find out if ancestral strains of E. amylovora were with or without the pEA29 plasmid. It is tempting to speculate that E. amylovora evolved by acquiring plasmids from other bacteria through a lateral gene transfer with subsequent selection-based deletion of the gene(s) for conjugal transfer. It is possible that the ancestral lines were plasmid-free and less aggressive on their host plants, and became more pathogenic by acquiring plasmid DNA from other microbial sources. Thus, plasmid-free strains may represent ancestral strains rather than a more recent lineage (Ochman et al., 2000).

Acknowledgements I wish to thank the Alexander von Humboldt Foundation in Bonn, Germany, for granting the Georg Forster fellowship and K. Geider at Julius Ku¨hn Institut for Plant Protection in Fruit Crops and Viticulture, Dossenheim, Germany, for providing plasmids and bacterial strains.

References Aldridge P, Metzger M, Geider K, 1997. Genetics of sorbitol metabolism in Erwinia amylovora and its influence on bacterial virulence. Molecular General Genetics 256, 611–9. Amuthan G, Mahadevanm A, 1993. Plasmid and pathogenicity in Xanthomonas oryzae pathovar oryzae, the bacterial blight pathogen of Oryzae sativa. Journal of Applied Microbiology 76, 529–38. Bellemann P, Jahn N, Theiler R, Geider K, 1990. Transposon mutagenesis of Erwinia amylovora. Acta Horticulture 272, 233–7. Bereswill S, Geider K, 1997. Characterization of the rcsB gene from Erwinia amylovora and its influence on exopolysaccharide synthesis and virulence of the fire blight pathogen. Journal of Bacteriology 179, 1354–61. Bereswill S, Pahl A, Bellemann P, Zeller W, Geider K, 1992. Sensitive and species-specific detection of Erwinia amylovora by PCR-analysis. Applied and Environmental Microbiology 58, 3522–6. Bereswill S, Bugert P, Bruchmu¨ller I, Geider K, 1995. Identification of the fire blight pathogen, Erwinia amylovora, by PCR assays



with chromosomal DNA. Applied and Environmental Microbiology 61, 2636–42. Bereswill S, Jock S, Bellemann P, Geider K, 1998. Identification of Erwinia amylovora by growth morphology on agar containing copper sulfate and by capsule staining with lectin. Plant Disease 82, 158–64. Bernhard F, Poetter K, Geider K, Coplin DL, 1990. The rcsA gene for Erwinia amylovora, identification, nucleotide sequence and regulation of exopolysaccharide biosynthesis. Molecular PlantMicrobe Interactions 3, 429–37. Bogs J, Bruchmu¨ller I, Erbar C, Geider K, 1998. Colonization of host plants by the fire blight pathogen Erwinia amylovora, marked with genes for bioluminescence and fluorescence. Phytopathology 88, 416–21. Bonn WG, van der Zwet T, 2000. Distribution and economic importance of fire blight. In: Vanneste JL, ed. Fire Blight: the Disease and its Causative Agent, Erwinia amylovora. Wallingford, UK: CABI International. Coplin DL, Majerczak DR, Zhang Y, Kim W-S, Jock S, Geider K, 2002. Identification of Pantoea stewartii subsp. sterwartii by PCR and strain differentiation by PFGE. Plant Disease 86, 304–11. Dellagi A, Brisset M-N, Paulin J-P, Expert D, 1998. Dual role of desferrioxamine in Erwinia amylovora pathogenicity. Molecular Plant-Microbe Interactions 11, 734–42. El Hassouni M, Chambost JP, Expert D, Van Gusegem F, Barras F, 1999. The minimal gene set member msrA, encoding peptide methionine sulfoxide reductase, is a virulence determinant of the plant pathogen Erwinia chrysanthemi. Proceedings of the National Academy of Sciences, USA 96, 887–92. Falkenstein H, Bellemann P, Walter S, Zeller W, Geider K, 1988. Identification of Erwinia amylovora, the fire blight pathogen, by colony hybridization with DNA from plasmid pEA29. Applied and Environmental Microbiology 54, 2798–802. Falkenstein H, Zeller W, Geider K, 1989. The 29 kb plasmid, common in strains of Erwinia amylovora, modulates development of fire blight symptoms. Journal of General Microbiology 135, 2643–50. Geider K, Auling G, Du Z, Jakovljevic V, Jock S, Vo¨lksch B, 2006. Erwinia tasmaniensis sp. nov, a non-phytopathogenic bacterium from apple and pear trees. International Journal of Systematic and Evolutionary Microbiology 56, 2937–43. Geier G, Geider K, 1993. Characterization and influence on virulence of the levansucrase gene from the fireblight pathogen Erwinia amylovora. Physiological and Molecular Plant Pathology 42, 387–404. Jock S, Geider K, 2004. Molecular differentiation of Erwinia amylovora strains from North America and of two Asian pear pathogens by analyses of PFGE patterns and hrpN genes. Environmental Microbiology 6, 480–90. Laurent J, Barny M-A, Kotoujansky A, Dufriche P, Vanneste JL, 1989. Characterization of a ubiquitous plasmid in Erwinia amylovora. Molecular Plant-Microbe Interactions 2, 160–4. Llop P, Bonaterra A, Penalver J, Lopez MM, 2000. Development of a highly sensitive nested-PCR procedure using a single closed tube for detection of Erwinia amylovora in asymptomatic plant material. Applied and Environmental Microbiology 66, 2071–8. Llop P, Donat V, Rodriguez M et al., 2006. An indigenous virulent strain of Erwinia amylovora lacking the ubiquitous plasmid pEA29. Phytopathology 96, 900–7.


261

McGhee GC, Jones AL, 2000. Complete nucleotide sequence of ubiquitous plasmid pEA29 from Erwinia amylovora strain E188: gene organization and intraspecies variation. Applied and Environmental Microbiology 66, 4897–907. McGhee GC, Sundin GW, 2007. Thiamin biosynthesis and its influence on exopolysaccharide production: a new component of virulence indentified on Erwinia amylovora plasmid pEA29. Acta Horticulturae 793, 271–7. McManus PS, Jones AL, 1994. Epidemiology and genetic analysis of streptomycin-resistant Erwinia amylovora from Michigan and evaluation of oxytetracycline for control. Phytopathology 84, 627–33. Maes M, Orye K, Bobev S et al., 2001. Influence of amylovoran production on virulence of Erwinia amylovora and different amylovoran structure in E. amylovora isolates from Rubus. European Journal of Plant Pathology 107, 839–44. Meletzus D, Bermpohl A, Dreier J, Eichenlaub R, 1993. Evidence for plasmid-encoded virulence factors in the phytopathogenic bacterium Clavibacter michiganensis subsp. michiganensis NCPPB382. Journal of Bacteriology 175, 2131–6. Mergaert J, Hauben I, Cnockaert MC, Swings J, 1999. Reclassification of non-pigmented Erwinia herbicola strains from trees as Erwinia billingiae sp. nov. International Journal of Systematic Bacteriology 49, 377–83. Metzger M, Bellemann P, Bugert P, Geider K, 1994. Genetics of galactose metabolism of Erwinia amylovora and its influence on polysaccharide synthesis and virulence of the fire blight pathogen. Journal of Bacteriology 176, 450–9. Nair GR, Liu Z, Binns AN, 2003. Reexamining the role of the accessory plasmid pAtC58 in the virulence of Agrobacterium tumefaciens strain C58. Plant Physiology 133, 989–99. Ochman H, Lawrence JG, Groisman EA, 2000. Lateral gene transfer and the nature of bacterial innovation. Nature 405, 229–304. Oh C-S, Beer SV, 2005. Molecular genetics of Erwinia amylovora involved in the development of fire blight. FEMS Microbiological Letters 253, 185–92. Rhim S-L, Vo¨lksch B, Gardan L et al., 1999. An Erwinia species, different from Erwinia amylovora, causes a necrotic disease of Asian pear trees. Plant Pathology 48, 514–20. Rico A, Preston GM, 2008. Pseudomonas syringae pv. tomato DC3000 uses constitutive and apoplast-induced nutrient assimilation pathways to catabolize nutrients that are abundant in the tomato apoplast. Molecular Plant-Microbe Interactions 21, 269–82. Schwyn B, Neilands JB, 1987. Universal chemical assay for the detection and determination of siderophores. Analytical Biochemistry 160, 47–56. Shaw JJ, Settles LG, Kado CI, 1988. Transposon Tn4431 mutagenesis of Xanthomonas campestris pv. campestris: characterization of a nonpathogenic mutant and cloning of a locus for pathogenicity. Molecular Plant-Microbe Interactions 1, 39–45. Venisse J-S, Barny M-A, Paulin J-P, Brisset M-N, 2003. Involvement of three pathogenicity factors of Erwinia amylovora in the oxidative stress associated with compatible interaction in pear. FEBS Letters 537, 198–202. Zhao Y, Blumer SL, Sundin GW, 2005. Identification of Erwinia amylovora genes induced during infection of immature pear tissue. Journal of Bacteriology 187, 8088–103.