Molecular analyses of Erwinia amylovora strains isolated in Russia, Poland, Slovenia and Austria describing further spread of fire blight in Europe

Microbiological Research 168 (2013) 447–454

Contents lists available at SciVerse ScienceDirect

Microbiological Research journal homepage: www.elsevier.com/locate/micres

Molecular analyses of Erwinia amylovora strains isolated in Russia, Poland, Slovenia and Austria describing further spread of fire blight in Europe Susanne Jock a , Annette Wensing a , Joanna Pulawska b , Nataliya Drenova c , Tanja Dreo d , Klaus Geider a,∗ a

Julius Kuehn Institute, Institute for Plant Protection in Fruit Crops and Viticulture, Schwabenheimer Str. 101, D-69221 Dossenheim, Germany Research Institute of Horticulture, Kostytucji 3 Maja 1/3, 96-100 Skierniewice, Poland All-Russian Center for Plant Quarantine, Pogranichnaya Str. 32, Ramensky Region, Moscow obl., Russia d National Institute of Biology, Vecna pot 111, SI-1000 Ljubljana, Slovenia b c

a r t i c l e

i n f o

Article history: Received 21 August 2012 Received in revised form 23 January 2013 Accepted 25 January 2013 Available online 6 April 2013 Keywords: Fire blight Molecular differentiation PFGE analysis Plasmids

a b s t r a c t Fire blight, a bacteriosis of apple and pear, was assayed with molecular tools to associate its origin in Russia, Slovenia and south-eastern Austria with neighboring countries. The identification of all investigated strains was confirmed by MALDI-TOF mass spectroscopy except one. Independent isolation was verified by the level of amylovoran synthesis and by the number of short sequence DNA repeats in plasmid pEA29. DNA of gently lysed E. amylovora strains from Russia, Slovenia, Austria, Hungary, Italy, Spain, Croatia, Poland, Central Europe and Iran was treated with restriction enzymes XbaI and SpeI to create typical banding patterns for PFGE analysis. The pattern Pt2 indicated that most Russian E. amylovora strains were related to strains from Turkey and Iran. Strains from Slovenia exhibited patterns Pt3 and Pt2, both present in the neighboring countries. Strains were also probed for the recently described plasmid pEI70 detected in Pt1 strains from Poland and in Pt3 strains from other countries. The distribution of pattern Pt3 suggests distribution of fire blight from Belgium and the Netherlands to Central Spain and Northern Italy and then north to Carinthia. The PFGE patterns indicate that trade of plants may have introduced fire blight into southern parts of Europe proceeded by sequential spread. © 2013 Elsevier GmbH. All rights reserved.

1. Introduction Fire blight is caused by the Gram-negative bacterium Erwinia amylovora and affects apple, pear, quince and other rosaceous plants. The disease was first described more than 200 years ago in North America (Denning 1794) and then was introduced into New Zealand from where it possibly spread to other parts of the world in the 20th century, such as Europe and the Mediterranean region (Bonn and van der Zwet 2000). E. amylovora can be identified by several methods. Detection of fire blight by PCR assays is in most cases reliable, and identification by comparing the protein profiles in MALDI-TOF mass spectroscopy is a fast approach (Sauer et al. 2008; Wensing et al. 2012). Most identification methods do not allow further sub-grouping of different strains, which would be useful in tracking the source of outbreaks for epidemiological purposes. American strains and strains isolated outside of North America can be distinguished by analysis of single nucleotide

∗ Corresponding author at: Julius Kühn Institut (JKI) für Pflanzenschutz, Schwabenheimer Str. 101, D-69221 Dossenheim, Germany. Tel.: +49 6221 86805 53; fax: +49 6221 86805 15. E-mail address: [email protected] (K. Geider). 0944-5013/$ – see front matter © 2013 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.micres.2013.01.008

polymorphisms (SNPs) in the genes galE, acrB and hrpA (Gehring and Geider 2012). The fact that non-American strains strictly share the SNP patterns indicates a single event for introduction of the disease by plant imports to New Zealand and subsequently to England and Egypt. On a more local scale, spread of fire blight to adjacent areas is possible by flower-visiting insects, rain and wind. Trade of infected plant material can also distribute the disease. The epidemiology of strains from a more localized region can be described by using pulsed-field gel electrophoresis (PFGE) restriction patterns to distinguish them. In general, Central European strains share the PFGE pattern Pt1, while the Pt3 pattern is dominant in Belgium and Northern France, and strains from the Eastern Mediterranean region exhibit mainly the Pt2 pattern. Especially the recent occurrence of the Pt3 type in Northern Italy and Central Spain was associated with plant trade (Jock et al. 2002). Other properties of E. amylovora may be also suited to distinguish strains, such as phage sensitivity (Müller et al. 2011), levan production (Bereswill et al. 1997) and the amount of EPS synthesis (Wang et al. 2010). Changes in the short sequence DNA repeat of the common plasmid pEA29 (Kim and Geider 1999) are also useful for strain differentiation. Here we use PFGE typing to follow the movement of fire blight disease from central to eastern Europe and from Greece westwards to Hungary. In the Russian Federation, fire blight was detected 2003

448

S. Jock et al. / Microbiological Research 168 (2013) 447–454

and first identified according to the EPPO Standard. Since 2007, the Russian quarantine service has been monitoring E. amylovora. Fire blight has now been reported in 11 regions of the European part of Russia. The first sporadic outbreaks of fire blight were seen in Slovenia in 2001 and 2002 (Knapiˇc et al. 2004; Dreo et al. 2006). Fire blight was discovered for the first time in Southwest Austria in 1993 on Cotoneaster salicifolius in Vorarlberg and later in the north of Austria (Keck et al. 2006), from where it then spread to other regions. Recent outbreaks were detected in Carinthia (southeast Austria). To follow this spread of fire blight, selected strains isolated in southern and eastern Europe were analyzed for their PFGE pattern types together with other properties that may be associated with the strains. MALDI-TOF mass spectroscopy, amylovoran production, and the presence of the 66 kb cryptic plasmid pEI70 were also used to analyze fire blight in parts of southern and eastern Europe. 2. Materials and methods 2.1. Bacterial strains The E. amylovora strains used in this study are listed in Table 1. They were stored frozen at −80 ◦ C in 10% glycerol and then grown in Luria–Bertani broth (LB) at 28 ◦ C. 2.2. PFGE analysis The genomic DNA from gently lysed cells was digested with restriction enzymes XbaI and SpeI as previously described (Jock et al. 2002). For lysis, the cells embedded in 1% agarose blocks, which were incubated three days at 50 ◦ C in lysis buffer, digested with 30 units enzyme, and then washed in phosphate/EDTA buffer. PFGE was performed with a CHEFDRIII apparatus (Bio-Rad) in a 1% agarose gel with a linear ramping of 5 V/cm for 1–25 s for 22 h at 14 ◦ C. 2.3. PCR assays qPCR analysis was done as previously described (Mohammadi et al. 2009; Wensing et al. 2012) using primers #101 (P29TF) #102b (P29TM), #103 (P29TR) (from plasmid pEA29); #107 (AR14819), #109 (AR14948c), #113 (AR14840Cy) (chromosomal amsK gene). The samples of the cultures at 1 d were diluted 100-fold in 0.1% Tween and the cells lysed 15 min at 65 ◦ C. For analysis of short sequence DNA repeats (SSR) numbers and plasmid pEI70, the primers are listed in Table 2. PCR fragments were sequenced commercially (Seqlab, Göttingen, Germany). 2.4. Amylovoran determinations Bacteria were grown in MM2C medium for 2 d at 28 ◦ C and amylovoran in the supernatant was measured with the CPC assay (Bellemann et al. 1994) in duplicate.

as matrix. Spectra were generated on a Bruker microflex machine with Biotyper standard settings and analyzed with Biotyper software 3.1. Scores above 2 indicate a match with reference spectra (Sauer et al. 2008). 3. Results 3.1. MALDI-TOF MS analysis for strain identification The strains of E. amylovora listed in Table 1 were analyzed by MALDI-TOF MS to verify their identification from previous assays. These strains were obtained over various years in Slovenia, Russia, Hungary, Austria, Italy, Croatia, Poland, Central Europe, Iran and other regions with fire blight. All but one of the strains was confirmed by MALDI-TOF to be E. amylovora with a score value of ca. 2.3 (Table 3) exceeding a threshold of 2 for reliable identification of species (Sauer et al. 2008). Isolate ErwKae28/07 from necrotic pear tissue in Carinthia, Austria, was identified as Brenneria quercina. This species may be a common saprophyte in necrotic wood of trees with fire blight. With this exception, all isolates showed a homogenous and specific protein pattern that allowed no further subtyping. 3.2. PFGE analysis with XbaI and SpeI E. amylovora strains isolated in the Eastern part of Europe with emphasis on southeast Austria, Poland, Russia and Slovenia were compared with those of strains from Central Europe, Hungary, Croatia and Northern Italy for their PFGE patterns generated by XbaI and SpeI digestion (Fig. 1 and Table 3). Previously described dominant patterns were confirmed: Pt1 for Germany and a set of 12 strains from Poland; Pt2 for Hungary, Croatia and Iran; and Pt3 for Northern Italy. The pattern distribution was split for other countries. For Austria three isolates from the southern part were Pt3 while two older isolates from the Vienna district were Pt1. In Croatia only Pt2 was found (Halupecki et al. 2006). Slovenia is surrounded by countries carrying Pt2 (Croatia) and Pt3 (Northern Italy) and both pattern types were observed during the 2007 outbreak. Based on the presence of the same PFGE patterns, even Carinthia could be a source of fire blight in parts of Slovenia. The Russian enclave of the Kaliningrad region was infected with Pt1, as were the surrounding areas of Poland, whereas other parts of Russia harbored Pt2 strains. Fire blight may have moved into these central regions from the south, possibly from Turkey and Iran. Strains from Turkey were previously shown to have the Pt2 pattern (Zhang and Geider 1997). The XbaI patterns clearly distinguished Pt1 and Pt2 strains. The results of this study also show that digestion with SpeI gives the same PFGE strain groupings as with XbaI. The SpeI digests produced a large DNA fragment with a lack of a smaller one in case of Pt3 strains. The XbaI pattern types Pt1 and Pt2 can therefore be clearly distinguished from Pt3 in SpeI digests (Fig. 1B, arrows). We have confirmed Pt3 XbaI pattern type with SpeI. These highly conserved PFGE patterns are typical for the genomes of E. amylovora strains.

2.5. MALDI-TOF mass spectroscopy 3.3. Occurrence of plasmid pEI70 Spectra generation and Biotyper analysis were performed as previously described (Sauer et al. 2008; Wensing et al. 2012). Briefly, cells from an overnight culture in LB medium with 1% glycerol were harvested, washed once, inactivated in ethanol and pelleted again. The dried pellet was resuspended in 70% formic acid for lysis. For MALDI-TOF MS analysis acetonitrile was added and debris removed by centrifugation. 1–2 ␮l of extracts were placed on an MSP 96 polished steel target using saturated alpha-cyano-4hydroxy cinnamic acid in 50% acetonitrile–2.5% trifluoroacetic acid

As previously reported (Jock et al. 2002) the PFGE pattern of strains from Belgium, the Netherlands, France and Central Spain were mostly Pt3. These and the other strains from Table 1 were analyzed for the presence of plasmid pEI70 by qPCR (Fig. 2). A signal was obtained with primers #721 and #722 for strains NIB Z 895 from Slovenia, EaKae8/07 and EaKae24/07 from Carinthia, Ea404VR, and OMP-BO1194/97, OMP-BO1792/97 from Italy, but not for NIB Z 972, NIB Z 981 and EaKae23/07, which were among the

S. Jock et al. / Microbiological Research 168 (2013) 447–454

449

Table 1 E. amylovora strains used in this study. Strains/origin Russia (this study) VNIIKR KBE1

Description of isolation

Source, Reference

Cydonia 2009, from Kabardino-Balkaria, Baksan region (North Caucasus, close to Georgia, East of Sochi, between Black Sea and Caspian Sea) Crataegus 2007, from Kaliningrad oblast, Baltijsk Cydonia 2010, from Volgograd oblast, Krasnoslobodsk (450 km up the Volga river from the Caspian Sea) Malus domestica 2010, from Voronezh oblast Novousmansk region (300 km south of Moscow) Pyrus communis 2010, from Tambov oblast, Michurinsk (300 km south of Moscow)

VNIIKR KE6 VNIIKR VGE3 VNIIKR VRE4 VNIIKR TE9 Slovenia (this study) NIB Z 884 M. domestica, 2007, Naklo (first detection of fire blight in 2001) NIB Z 895 Cydonia oblonga, 2007, Lukovica NIB Z 972 M. domestica, 2007, Horjul NIB Z 981 M. domestica, 2007, Domˇzale NIB Z 997 M. domestica, 2007, Kriˇzevci Austria Ea296/93 C. salicifolus, Vienna, 1993 Ea674/94 P. communis, Vienna,1994 EaKae8/07 Necrotic pear wood, Carinthia, 2007 EaKae23/07 Necrotic pear wood, Carinthia, 2007 EaKae24/07 Necrotic pear wood, Carinthia, 2007 ErwKae28/07 From necrotic pear, Carinthia, 2007, is B. quercina (MALDI-TOF MS score 2.046) Belgium and the Netherlands DGBBC 350 Malus tsonoski, tree nursery, Oosterzele DGBBC 351 P. communis, Beech Hill’, tree nursery, Oosterzele LMG 1880 P. communis, 1979 LMG 1884 Cotoneaster salicifolius, 1980 LMG 1893 P. communis, 1980 M. domestica OVG038 PD576 Crataegus sp., the Netherlands, 1985 PD3217 Salix sp., the Netherlands, 1998 Croatia EaCr8DJ M. domestica, Djakovo, Croatia 1998 EaCr20/05 P. communis; Èakovec EaCr12/05 M. domestica, Gloster; Nedelisce Germany Ea7/74 Cotoneaster sp., Northern Germany 1974 Ea1/79 DSM 17948, M. domestica, Germany, 1979 Ea63/05 Pfullingen, Germany, Pyracantha sp., 2005; without plasmid pEA29 Cotoneaster sp., Rottweil, Germany 2006 EaRW1/06 Hungary EaH2 M. domestica, 1995, Kecskemet EaH895 M. domestica, ‘Golden Dilicious’, from leaf vein, Nyarlorinc, 1996 EaH909 M. domestica, ‘Jonathan’, from shoot, Zakanyszek, 1996 EaH910 C. oblonga, from shoot, Mohacs, 1996 Italy Ea404VR P. communis, district of Verona OMP-BO1194/97 P. communis, district of Ferrara, 1997 OMP-BO1792/97 P. communis, district of Modena, 1997 Iran EaIrn2 Pear leaf petiole, Shahriar Karaj, Tehran province, 2001 EaIrn37 P. communis, Tabriz, 2001; without pEA29 Poland (this study) ˛ EaPL2 Crataegus sp., O´swiecim, 2000 ˛ EaPL600 Crataegus sp., Kepno, 1993 ˛ EaPL601 P. communis, Kepno, 1993 ˛ EaPL603 Crataegus sp., Kepno, 1994 ˙ EaPL606 Crataegus sp., Elzbietów, 1994 ˛ EaPL665 P. communis, Kepno, 1993 Crataegus sp., Opole, 1996 EaPL666 EaPL609a P. communis, Łowicz, 2003 EaPL619a P. communis, Biała Rawska, 2007 EaPL625a P. communis, Grójec, 2007 EaPL367/96 Pyrancantha sp., 1996 EaPL692/95 Sorbus sp., 1995 Spain IVIA 1603 Cotoneaster sp., Segovia, 1996 IVIA 1614-2 Pyracantha sp., Segovia, 1996 IVIA 1899-21 C. oblonga, Guadalajara, 1998 Strains from other countries T Type strain, P. communis, England, 1959 CFBP 1232 CFBP 1430 Crataegus sp., Lille, France, 1972 Ea4/82 P. communis, Egypt, 1982 T90 Turkey, 1990 Ea775 NCPPB775; Crataegus sp.; England, 1959 OR6 P. communis, Oregon, USA, 2007 JLVNZ04 Hawke’s bay, New Zealand, 1994 PFB15 Prunus sp., USA, before 1995 Tp3 P. communis, Toronto, 2001

Jock et al. (2003) Bereswill et al. (1997) This study This study This study This study Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Bereswill et al. (1997) Jock et al. (2002) Halupecki et al. (2006) Halupecki et al. (2006) Halupecki et al. (2006) Falkenstein et al. (1988) Falkenstein et al. (1988) Mohammadi et al. (2009) K. Geider M. Hevesi J. Nemeth A. N. Kovacs A. N. Kovacs Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Mohammadi et al. (2009) Mohammadi et al. (2009)

P. Sobiczewski P. Sobiczewski Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Jock et al. (2002) Falkenstein et al. (1988) Zhang and Geider (1997) Zhang and Geider (1997) V. Stockwell Kim and Geider (1999) Kim and Geider (1999) Jock and Geider (2004)

BCCM/LMG, Belgian co-ordinated collections of micro-organisms; CFBP, Collection Franc¸aise de Bactéries Phytopathogènes; DSMZ, Deutsche Sammlung von Mikroorganismen und Zellkulturen; NIB Z, bacterial culture collection of the National Institute of Biology, Ljubljana, Slovenia; VNIIKR, Russian Centre for Plant Quarantine.

450

S. Jock et al. / Microbiological Research 168 (2013) 447–454

Table 2 Novel PCR primers used in this study. Primer

Other name

Sequence

Sequencing of SSRs in plasmid pEA29 Ea29SSR25638 GGCCTATGCCGTCTCAGAAT #729 #730 Ea29SSR26367c GCTGGTAGCGATGTTGATGA cPCR for detection and sequencing part of plasmid pEI70 (the numbers indicate primer positions, Genbank accession number CP002951) PEI70-52130 ACCTGAAGTCGCACCGGATA #719 #720 PEI70-52924c TGGCGTAGCTGGATTAGTCG qPCR, detection of plasmid pEI70 PEI70-53061Q AGAAGTTAGCGCGAGATGGA #721 PEI70-53226Qc TAGTCCTGCGTGGACGACTA #722

Fig. 1. PFGE analysis. (A) Digest with enzyme XbaI. Lane 1: CFBP1232 (England, Pt1); 2: Ea4/82 (Egypt, Pt2); 3: CFBP1430 (France, Pt3); 4: NIB Z 884 (Slovenia, Pt2); 5: NIB Z 981 (Slovenia, Pt3); 6: VNIIKR KE6 (Kaliningrad; Russia; Pt1); 7: VNIIKR TE9 (Michurinsk, Russia; Pt2). M, marker of lambda oligomers; asterisks, band positions indicating Pt2. (B) Digests with enzyme SpeI. Lane 1: CFBP1232T (England, Pt1); 2: Ea4/82 (Egypt, Pt2); 3: CFBP1430 (France, Pt3); 4: NIB Z 981 (Slovenia, Pt3); 5: VNIIKR KE6 (Kaliningrad, Russia; Pt1); 6: EaKae8/07 (Carinthia, Austria; Pt3); 7: Ea404VR (Verona, Italy; Pt3); 8: NIB Z 895 (Slovenia; Pt3). M, marker of lambda oligomers in kb; arrow with dot: typical band for Pt3; arrow with diamond: missing band for Pt3.

Pt3 strains from these countries. The positive qPCR data were confirmed by cPCR with primers #719/#720. The nucleotide sequences of the amplified fragments were identical. To determine if plasmid pEI70 is associated with Pt3 strains as initially observed, we assayed 12 E. amylovora strains from Poland (Table 3), most were identified before as carriers of the plasmid (Llop et al. 2011). In SpeI digests all strains from Poland displayed the pattern type Pt1, i.e. pEI70 can be carried by strains with PFGE pattern types Pt1 and Pt3. Most of the Pt3 strains from Belgium, the Netherlands, France, and Central Spain were positive for pEI70. None of the tested E. amylovora strains from North America, nor the single strain from New Zealand (JLVNZ04) used in this study, carried the plasmid. In conclusion, the PFGE patterns and the presence of pEI70 plasmid, together with limited information on plant trade indicate that trade from Belgium/the Netherlands to Central Spain and Northern Italy was a possible route for long distance distribution of fire blight.

9000 8000 7000

RFU

6000 5000 4000 3000

3.4. DNA repeat numbers of plasmid pEA29

A

2000 1000

B

0 0

10

20

30

40

Cycles Fig. 2. Screening with qPCR for presence of pEI70 using primers #721 and #722. (A) Positive signal for pEI70 from strains: DGBBC 350, LMG 1880, LMG 1893, PD576, IVIA 1603, Sp1899-2, EaKae24/07. (B) No signal (no pEI70): OVG038, EaKae23/07.

The E. amylovora strains from Russia, Slovenia and Austria were investigated for the number of short-sequence DNA repeats (SSR) in the 1 kb PstI fragment of plasmid pEA29. This plasmid region was amplified with primers #721/#722 and sequenced (Table 3). The strains from Slovenia had repeat numbers from 6 to 9. The strains from Russia showed lower repeat numbers from 4 to 6. Rarely observed, Russian strain VNIIKR KBE1 had a different nucleotide sequence in the right part of the amplified PCR fragment, whereas the rest of the fragment matched with the expected sequence.

S. Jock et al. / Microbiological Research 168 (2013) 447–454

451

Table 3 MALDI-TOF MS identification of E. amylovora strains, presence of plasmid pEI70, number of SSRs in pEA29, and PFGE Pt types. Strain/origin Russia VNIIKR KBE1 VNIIKR KE6 VNIIKR VGE3 VNIIKR VRE4 VNIIKR TE9 Slovenia NIB Z 884 NIB Z 895 NIB Z 972 NIB Z 981 NIB Z 997 Austria EaKae8/07 EaKae23/07 EaKae24/07 Ea296/93 Ea674/94 Belgium, the Netherlands DGBBC 350 DGBBC 351 LMG 1880 LMG 1884 LMG 1893 OVG038 PD576 PD3217 Croatia EaCr8DJ EaCr12/05 EaCr20/05 Germany Ea7/74 Ea1/79 Ea63/05 (no pEA29) EaRW1/06 Hungary EaH2 EaH895 EaH909 EaH910 Italy Ea404VR OMP-BO1194/97 OMP-BO1792/97 Iran EaIrn2 EaIrn37 (no pEA29) Poland EaPL2 EaPL600 EaPL601 EaPL603 EaPL606 EaPL665 EaPL666 EaPL609a EaPL619a EaPL625a EaPL367/96 EaPL692/95 Spain IVIA 1603 IVIA 1614-2 IVIA 1899-2 Other strains CFBP 1232T CFBP 1430 Ea4/82 T90 Ea775 OR6 JLVNZ04 PFB15 Tp3 a

MALDI-TOF MS score

pEI70

SSR

PFGE pattern; reference

2.285 2.274 2.320 2.297 2.286

No No No No No

a

5 6 6 4

Pt2, this study Pt1, this study Pt2, this study Pt2, this study Pt2, this study

2.394 2.449 2.335 2.368 2.258

No Yes No No No

6 9 6 7 8

Pt2, this study Pt3, this study Pt3, this study Pt3, this study Pt2, this study

2.354 2.308 2.262 2.324 2.360

Yes No Yes No No

7 6 6 14 13

Pt3, this study Pt3, this study Pt3, this study Pt1, this study Pt1, this study

2.338 2.243 2.369 2.400 2.374 2.343 2.353 2.335

Yes Yes Yes Yes Yes No Yes Yes

Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3s (Jock et al. 2002) Pt3 (Jock et al. 2002)

2.271 2.277 2.356

No No No

Pt2, this study and Halupecki et al. (2006) Pt2, this study and Halupecki et al. (2006) Pt2, this study and Halupecki et al. (2006)

2.172 2.352 2.254 2.203

No No No No

2.258 2.360 2.354 2.403

No No No No

Pt2, this study Pt2, this study Pt2, this study Pt2, this study

2.318 2.323 2.407

Yes Yes Yes

Pt3, this study and Jock et al. (2002) Pt3, this study and Jock et al. (2002) Pt3, this study and Jock et al. (2002)

2.324 2.318

No No

Pt2, this study Pt2, this study

2.435 2.355 2.262 2.322 2.333 2.280 2.371 2.372 2.321 2.350 2.334 2.277

Yes No Yes Yes Yes Yes Yes Yes Yes Yes No No

Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study Pt1, this study

2.386 2.449 2.359

Yes Yes Yes

Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002) Pt3 (Jock et al. 2002)

2.407 2.512 2.227 2.207 2.153 2.289 2.184 2.281 2.177

No No No No No No No No

6 6

5 6 5 7

9

Changes in pEA29, no SSR detected. A MALDI-TOF MS score > 2 identifies E. amylovora; SSR, short sequence DNA-repeats.

Pt1 (Zhang et al. 1998) Pt1 (Zhang et al. 1998) Pt1, this study

Pt1 (Jock et al. 2002) Pt3, this study and Jock et al. (2002) Pt2, this study and Jock et al. (2002) Pt2 (Zhang and Geider 1997) Pt4 (Jock et al. 2002)

Other (Kim and Geider 1999) Other (Jock and Geider 2004)

452

S. Jock et al. / Microbiological Research 168 (2013) 447–454 8

7

6

A 600

5

4

3

2

1

CF BP

12 E 32 CF a4/8 BP 2 1 NI 430 BZ NI 884 BZ NI 895 BZ NI 972 BZ 98 VN NIBZ 1 IIK 9 R 97 VN KB E I I K VN R 1 IIK KE VN R V 6 IIK GE R 3 VN VR IIK E4 R TE Ea 9 Ea H2 H8 Ea 95 H9 Ea 09 Ea H91 K 0 Ea ae8 Ka /07 e2 Ea 8/07 29 Ea 6/93 67 OM Ea 4/94 40 P OM -BO 4VR P- 119 BO 4/9 17 7 9 Ea 2/97 Cr Ea 8D J Cr Ea 20/0 Cr 5 12 /0 Ea 5 Ir E n2 Ea aIrn PL 37 Ea 367 PL /96 69 2/ Ea 9 5 1 Ea /79 63 /05

0

Fig. 3. Amylovoran synthesis. The cultures were grown in MM2C liquid medium for 2 d. After addition of CPC, the turbidity was measured at 600 nm.

Surprisingly, primer #722 annealed at a site to amplify a fragment of ca. 900 bp. On the other hand, no homology was found for the right part of the fragment in BLAST searches to any sequence in nucleotide data bases possibly indicating a recent genomic rearrangement. Different SSR numbers indicate that non-identical E. amylovora strains occur, even in limited regions such as for Slovenia. Also strains from Carinthia differed in their SSR number. 3.5. Amylovoran synthesis and growth in minimal medium The amylovoran production in MM2 minimal medium was correlated with bacterial growth determined by qPCR. The Taqman primers #102b (FAM) from plasmid pEA29 and for confirmation #113 (Cy5) from the chromosomal ams region were simultaneously applied to measure growth after 1 d of incubation. Use of ams primers was a precaution to record strains with a change in plasmid pEA29 such as VNIIKR KBE1, which may not produce a signal. The Ct values for all Tween lysates were almost identical and matched with a lysate of E. amylovora strain Ea1/79 at a cell density of 1 × 109 CFU/ml. Different amounts of amylovoran synthesis between strains therefore cannot be attributed to differences in cell growth. The capsular EPS amylovoran is strictly required for full pathogenicity (Bellemann and Geider 1992). Among the investigated strains, CFBP 1430 from France and Ea296/93 from Austria were high amylovoran producers (Fig. 3). Among the strains isolated outside Slovenia, Russia and Austria, Ea4/82 from Egypt, EaIrn2 and EaIrn37 from Iran, EaH2 and EaH910 from Hungary, EaPL367/96 from Poland synthesized small amounts of amylovoran. The level of most Russian and Slovenian strains was intermediate and VNIIKR VRE4 and VNIIKR TE9 from Russia showed a more reduced amylovoran synthesis than the others. 3.6. Occurrence and characterization of fire blight in Russia, Poland, Slovenia and southeast Austria On the European mainland, fire blight was detected 1966/1967 in the Netherlands and on the Baltic coast of Poland (Bonn and van der Zwet 2000). At present, fire blight is observed with variable intensity in many apple and pear growing regions, mostly in

the central and south part of Poland. Besides fruit trees, fire blight was also found on ornamental plants such as Crataegus sp., Sorbus sp., Amelanchier sp. The analysis of E. amylovora isolates originating from different geographical regions of Poland and different host plants showed a high genetic homogeneity but revealed differences in virulence (Puławska et al. 2006). In the recent years, several new loci of fire blight, possibly connected with imported plant material was observed. Nevertheless, the strains from Poland investigated here showed the PFGE pattern Pt1, whereas Belgium/the Netherlands also harbor Pt3 strains with and without plasmid pEI70. Four clusters of fire blight infection exist in Russia. In the Kaliningrad oblast, a western enclave, fire blight was widespread 2003 on Crataegus sp. and was also detected on Malus, Pyrus, Amelanchier and Prunus spp. In contrast to other Russian clusters, the PFGE pattern observed here was Pt1. The second cluster in Russia is in the Central Black Earth Area with remote parts of the Voronezh region (from 2007) with affected old apple and pear trees. There are also locally infested areas in the Tambov region (2007, Michurinsk), and fire blight was found in the Belgorod oblast (2008) on grafted pear scions from Ukraine, and later in two remote locations of Lipetsk oblast (2011, 2012) on pear and apple trees in private and commercial gardens. The third cluster is in the Stavropol territory (Caucasian Mineral Waters) and the North Caucasus (Kabardino-Balkaria and Karachaevo-Cherkessia) with pear and quince production. Severe quarantine measures (2008–2010) attempted to reduce fire blight on Malus, Pyrus and Cydonia in private gardens and in orchards. In the fourth cluster, the Lower Volga area which has a continental climate with winters as cold as −40 ◦ C, a short spring and a very hot and dry summer (Samara oblast; 2008, 2011). Several old apple orchards in remote areas were eradicated, but outbreaks were detected in young plantations after two years latent infection in two farms (2012). In the Saratov oblast (Khvalynsk, 2009, 2011) fire blight was detected in apple and pear orchards and on wild Malus, Pyrus and Crataegus spp. Samara and Saratov oblasts are both infected with Pseudomonas syringae, which often resembles E. amylovora infections of pear blossoms and shoots. Near Volgograd (2009) fire blight affected Cydonia and Crataegus in private and botanical gardens with almost 100 years old collections in part from America. Quince showed symptoms

S. Jock et al. / Microbiological Research 168 (2013) 447–454

in June and mature apple trees carried single small shoot necrosis, and young apple trees can occasionally have typical symptoms. The impact of fire blight in this area is variable. We have mostly analyzed strains from the central southern parts of Russia. The PFGE pattern type Pt2 was also observed in the south in Turkey and in Iran. Slovenia was found to be free of fire blight until 2001. A low incidence of the disease was reported in 2001 and 2002 in the Gorenjska region. Despite strict phytosanitary measures implemented after discovery of the first focus, the bacterium spread in 2003 (Knapiˇc et al. 2004). Further spread of the disease and larger outbreaks were later recorded in years 2003 and 2007. Apart from these larger outbreaks, fire blight currently appears only sporadically in Slovenia. All Slovenian isolates were identified and confirmed before as E. amylovora using isolation and morphology observation on NSA and King’s B media, agglutination test, PCR (Pirc et al. 2009) and pathogenicity tests with immature pear fruits. Isolates for this study were selected on their variable number of tandem repeats profiles (T. Dreo et al., unpublished data). The analyzed strains were isolated in orchards near the Slovenian capital of Ljubljana within 100 km distance. Their PFGE pattern types are either Pt2 or Pt3. Plasmid pEI70 was not found in Pt2 strains from Slovenia. Fire blight was first detected in Austria in 1993. Since its first occurrence in Vorarlberg (southwest), close to the German border, the disease has spread in the north of Austria to other parts of the country (Keck et al. 2006). During the last disastrous outbreak in Vorarlberg in 2007, a substantial amount of the affected production area had to be cleared. The disease has more recently been observed in Carinthia (southeast) at altitudes as high as 1500 m (L.-M. Bartosik, personal communication). The E. amylovora strains from Carinthia (EaKae) showed the PFGE pattern type Pt3 and two of them carried plasmid pEI70. 4. Discussion Erwinia amylovora causes fire blight on Rosaceous plants and is a highly homogeneous species. We characterized a number of strains from fire blight outbreaks in Russia, Poland, Slovenia and southeast Austria (Carinthia) for their molecular properties and for the level of amylovoran synthesis. The identification of these strains as E. amylovora was verified by both characteristic PFGE patterns and MALDI-TOF MS. Screening of the strains in Table 1 is a broad identification attempt with MALDI-TOF MS analysis for E. amylovora from European sources and isolates from other regions. Only one strain was misidentified before, and it turned out to be Brenneria quercina. The E. amylovora strains differed not only in their PFGE patterns but also in the presence of plasmid pEI70. This plasmid is stably maintained in E. amylovora, and autonomously transferred to other E. amylovora strains (Llop et al. 2011). The number of SSRs of plasmid pEA29 (Kim and Geider 1999) and the amount of amylovoran synthesized, the hypersensitive response on tobacco, swarming on agar, and virulence on apple have been associated with strain signatures (Wang et al. 2010). These and other physiological properties were described before as strain characteristics (Halupecki et al. 2006). Among the Russian, Slovenian and Austrian strains, Ea674/94 from Northern Austria, NIB Z 895 from Slovenia and the Russian strain VNIIKR KE6 from the Baltic area were high amylovoran producers, whereas EaPL367/96 from adjacent Poland and EaH2 from adjacent Hungary were low producers. Most E. amylovora strains possess the common plasmid pEA29. It carries genes encoding thiamine metabolism (McGhee and Jones 2000), which may enhance growth of the pathogen in host plant tissue. A few, still virulent strains without plasmid pEA29, possibly from a spontaneous curing, have been described (Llop et al. 2006; Mohammadi et al. 2009). A frequent variation in the 1 kb PstI

453

fragment of plasmid pEA29 is a change in the repeat number of the 8 bp sequence “ATTACAGA” (Kim and Geider 1999). Natural E. amylovora populations are assumed to carry the same repeat number for all strains in a narrow geographical area, although isolates from adjacent plants in the same garden with fire blight can carry different SSR numbers (Jock et al. 2003). Other studies (Ruppitsch et al. 2004) consider SSR numbers to be characteristic for individual E. amylovora strains under lab conditions. In this study, we distinguished most of the isolates from Russia, Slovenia and southeast Austria (Carinthia) by SSRs. The reason for the mutation in plasmid pEA29 in VNIIKR KBE1 can be a recent recombination event adjacent to the SSR sequence. In addition to plasmid pEA29, E. amylovora may carry other plasmids. A broad investigation of European strains describes the occurrence of plasmid pEI70 and reports that it is present in the majority of E. amylovora strains from Italy and Slovenia (Llop et al. 2011). The presence of either plasmid can increase aggressiveness of strains, and pEI70 shows autonomous transfer. A dominance of positive strains in some countries may indicate their more efficient spread compared to plasmid free strains. For the E. amylovora strains tested here, pEI70 was not found yet in Pt2 strains. The absence of pEI70 in many Pt1 and Pt3 strains may indicate an evolutionary difference. A clear differentiation of E. amylovora isolates was achieved with PFGE analysis, a method that is difficult to use for mass screening but is very powerful because it compares whole genomes, rather than individual markers. Restriction analysis with XbaI has been successfully used for classification of European strains (Jock et al. 2002). Four main pattern types were observed among strains from Europe. Pt1 was found in Central Europe, Pt2 in Egypt and the Eastern Mediterranean region to Hungary, Pt3 in Northern Italy, Belgium, Northern France and Central Spain and Pt4 was found in Western France and Northern Spain. The comparison of XbaI restriction sites in the total genomic sequences of the American Pt1 strain Ea273 (accession no. FN666575, Sebaihia et al. 2010) with those in the genomic sequence of the French Pt3 strain CFBP 1430 (Jock et al. 2002) (accession number FN434113, Smits et al. 2010) revealed that different XbaI banding patterns can be explained by large inversions in the genomes. PCR analysis indicated that the respective XbaI and SpeI sites are still present (M. Gernold and K. Geider, unpublished data). In this study, we successfully applied pattern discrimination to E. amylovora strains isolated in Russia, Slovenia and southeast Austria (Carinthia). The pattern type in Russia was mostly Pt2, which was previously found in Turkey (Zhang and Geider 1997) and more recently in Iran. An exception was a strain from Kaliningrad, which exhibited the pattern Pt1, previously found in adjacent Poland (Jock et al. 2002). Therefore, we assume spread of fire blight from the south into western-central Russia. Slovenia was accordingly infiltrated from Croatia in the east, and from Northern Italy in the south. Pt3 in Carinthia can also be explained by spread of fire blight from the adjacent part of Northern Italy. This study expands the data on PFGE patterns of E. amylovora isolates present in Europe to strains from Russia, Slovenia and southeast Austria (Carinthia) (Fig. 4). The situation in Northern Italy is divergent: Pt3 is dominant but in rare cases Pt1 appears to have been introduced into the area of Bolzano from Austria (Jock et al. 2002). In a recent survey of PFGE patterns for E. amylovora strains from Serbia (Ivanovic´ et al. 2012), Pt2 was dominant, although Pt3 was also detected. The occurrence of additional patterns (Pt6, 7, 8, 9) may indicate a tendency of E. amylovora to undergo genomic rearrangements. PFGE-analysis of E. amylovora from Bulgaria also showed a dominance of Pt2 with some Pt1 strains and new patterns (Atanasova et al. 2012). Pt5 was described before in strains from Israel and Bulgaria together with patterns related to Pt3 from Belgium, the Netherlands and France (Jock et al. 2002).

454

S. Jock et al. / Microbiological Research 168 (2013) 447–454

The Netherlands Belgium (Northern) France

Germany

Russia

(Kaliningrad) Pt1

Poland

Pt2

Pt1 (Northern)

Pt3

(Carinthia) Slovenia Pt3

(Northern) (Central)

Iran

Austria

Pt2

Turkey Balkan states

Italy

Spain

Pt2

Pt2

Egypt

Fig. 4. A scheme for possible distribution of fire blight based on the PFGE patterns in Central Europe, Russia and the Eastern Mediterranean region with Iran.

The PFGE patterns of North American strains are heterogeneous (Jock and Geider 2004). Nevertheless, these strains can be classified by the SNP pattern of three genes divergent from that of all non-North American Isolates (Gehring and Geider 2012). This distribution might indicate, that spread of fire blight was caused by a rare escape of E. amylovora from North America to New Zealand from where it may have spread to Europe and the Mediterranean region by commercial trade. SNP analysis may become a tool also describing spread of fire blight in narrow areas supporting the data obtained by PFGE analysis. Acknowledgements We thank Marja-Liisa Bartosik for sending samples from symptomatic pear trees in Carinthia, Austria, and S. Zimmermann in the Hygiene Institute Heidelberg for providing access to Bruker Microflex-Biotyper analysis as well as David L. Coplin for valuable comments on the manuscript. The Slovenian isolates were obtained in the frame of official fire blight monitoring by an organization of the Slovenian Phytosanitary Administration and Phytosanitary Inspection. References Atanasova I, Urshev Z, Hristova P, Bogatzevska N, Moncheva P. Characterization of Erwinia amylovora strains from Bulgaria by pulsed-field gel electrophoresis. Z Naturforsch C 2012;67:187–94. Bellemann P, Geider K. Localization of transposon insertions in pathogenicity mutants of Erwinia amylovora and their biochemical-characterization. J Gen Microbiol 1992;138:931–40. Bellemann P, Bereswill S, Berger S, Geider K. Visualization of capsule formation by Erwinia amylovora and assays to determine amylovoran synthesis. Int J Biol Macromol 1994;16:290–6. Bereswill S, Jock S, Aldridge P, Janse JD, Geider K. Molecular characterization of natural Erwinia amylovora strains deficient in levan synthesis. Physiol Mol Plant Pathol 1997;51:215–25. Bonn GW, van der Zwet T. Distribution and economic importance of fire blight. In: Vanneste JL, editor. Fire blight the disease and its causative agent, Erwinia amylovora. Wallingford, UK: CABI Publishing; 2000. p. 37–54. Denning W. On the decay of apple trees. NY Soc Prom Agric Arts Manuf Trans 1794;2:219–22. Dreo T, Zupanˇciˇc M, Demsar T, Ravnikar M. First outbreaks of fire blight in Slovenia. Acta Horticult 2006;704:37–42.

Falkenstein H, Bellemann P, Walter S, Zeller W, Geider K. Identification of Erwinia amylovora, the fireblight pathogen, by colony hybridization with DNA from plasmid pEA29. Appl Environ Microbiol 1988;54:2798–802. Gehring I, Geider K. Differentiation of Erwinia amylovora and E. pyrifoliae strains with single nucleotide polymorphisms and by synthesis of dihydrophenylalanine. Curr Microbiol 2012;65:73–84. Halupecki E, Bazzi C, Jock S, Geider K, Dermic D, Cvjetkovic B. Characterization of Erwinia amylovora strains from Croatia. Eur J Plant Pathol 2006;114:435–40. Ivanovic´ M, Obradovic´ A, Gaˇsic´ K, Minsavage G, Dickstein E, Jones J. Exploring diversity of Erwinia amylovora population in Serbia by conventional and automated techniques and detection of new PFGE patterns. Eur J Plant Pathol 2012;133:715–27. Jock S, Geider K. Molecular differentiation of Erwinia amylovora strains from North America and of two Asian pear pathogens by analyses of PFGE patterns and hrpN genes. Environ Microbiol 2004;6:480–90. Jock S, Donat V, Lopez MM, Bazzi C, Geider K. Following spread of fire blight in Western, Central and Southern Europe by molecular differentiation of Erwinia amylovora strains with PFGE analysis. Environ Microbiol 2002;4:106–14. Jock S, Jacob T, Kim WS, Hildebrand M, Vosberg HP, Geider K. Instability of short-sequence DNA repeats of pear pathogenic Erwinia strains from Japan and Erwinia amylovora fruit tree and raspberry strains. Mol Genet Genomics 2003;268:739–49. Keck M, Richter S, Loncaric L, Ruppitsch W, Pechhacker H, Moosbeckhofer R, et al. Ten years of fire blight in Austria: survey and control measures. Acta Horticult 2006;704:43–50. Kim WS, Geider K. Analysis of variable short-sequence DNA repeats on the 29 kb plasmid of Erwinia amylovora strains. Eur J Plant Pathol 1999;105:703–13. Knapiˇc V, Potonik A, Skerlavaj V, Brecl A. First outbreaks of fireblight in Slovenia. EPPO Bull 2004;34:351–6. Llop P, Donat V, Rodriguez M, Cabrefiga J, Ruz L, Palomo JL, et al. An indigenous virulent strain of Erwinia amylovora lacking the ubiquitous plasmid pEA29. Phytopathology 2006;96:900–7. Llop P, Cabrefiga J, Smits TH, Dreo T, Barbe S, Pulawska J, et al. Erwinia amylovora novel plasmid pEI70: complete sequence, biogeography, and role in aggressiveness in the fire blight phytopathogen. PLoS One 2011;6:e28651. McGhee GC, Jones AL. Complete nucleotide sequence of ubiquitous plasmid pEA29 from Erwinia amylovora strain Ea88: gene organization and intraspecies variation. Appl Environ Microbiol 2000;66:4897–907. Mohammadi M, Moltmann E, Zeller W, Geider K. Characterisation of naturally occurring Erwinia amylovora strains lacking the common plasmid pEA29 and their detection with real-time PCR. Eur J Plant Pathol 2009;124:293–302. Müller I, Lurz R, Kube M, Quedenau C, Jelkmann W, Geider K. Molecular and physiological properties of bacteriophages from North America and Germany affecting the fire blight pathogen Erwinia amylovora. Microb Biotechnol 2011;4:735–45. Pirc M, Ravnikar M, Tomlinson J, Dreo T. Improved fireblight diagnostics using quantitative real-time PCR detection of Erwinia amylovora chromosomal DNA. Plant Pathol 2009;58:872–81. Puławska J, Kielak K, Sobiczewski P. Phenotypic and genetic diversity of selected Erwinia amylovora strains from Poland. Acta Horticult 2006;704:439–44. Ruppitsch W, Stoger AR, Keck M. Stability of short sequence repeats and their application for the characterization of Erwinia amylovora strains. FEMS Microbiol Lett 2004;234:1–8. Sauer S, Freiwald A, Maier T, Kube M, Reinhardt R, Kostrzewa M, et al. Classification and identification of bacteria by mass spectrometry and computational analysis. PLoS One 2008;3:e2843. Sebaihia M, Bocsanczy AM, Biehl BS, Quail MA, Perna NT, Glasner JD, et al. Complete genome sequence of the plant pathogen Erwinia amylovora strain ATCC 49946. J Bacteriol 2010;192:2020–1. Smits THM, Rezzonico F, Kamber T, Blom J, Goesmann A, Frey JE, et al. Complete genome sequence of the fire blight pathogen Erwinia amylovora CFBP 1430 and comparison to other Erwinia spp. Mol Plant-Microbe Interact 2010;23:384–93. Wang D, Korban SS, Zhao Y. Molecular signature of differential virulence in natural isolates of Erwinia amylovora. Phytopathology 2010;100:192–8. Wensing A, Gernold M, Geider K. Detection of Erwinia species from the apple and pear flora by mass spectroscopy of whole cells and with novel PCR primers. J Appl Microbiol 2012;112:147–58. Zhang Y, Geider K. Differentiation of Erwinia amylovora strains by pulsed-field gel electrophoresis. Appl Environ Microbiol 1997;63:4421–6. Zhang Y, Merighi M, Bazzi C, Geider K. Genomic analysis by pulsed-field gel electrophoresis of Erwinia amylovora strains from the Mediterranean region including Italy. J Plant Pathol 1998;80:225–32.