Volume 67, Issue 4 p. 957-970
Original Article
Free Access

Pathogenicity and phylogenetic analysis of Clavibacter michiganensis strains associated with tomato plants in Iran

E. Osdaghi

Corresponding Author

E. Osdaghi

Department of Plant Protection, College of Agriculture, Shiraz University, Shiraz, 71441-65186 Iran

E-mail: [email protected]Search for more papers by this author
M. Ansari

M. Ansari

Department of Plant Protection, College of Agriculture, Shiraz University, Shiraz, 71441-65186 Iran

Search for more papers by this author
S. M. Taghavi

S. M. Taghavi

Department of Plant Protection, College of Agriculture, Shiraz University, Shiraz, 71441-65186 Iran

Search for more papers by this author
S. Zarei

S. Zarei

Department of Plant Protection, College of Agriculture, Shiraz University, Shiraz, 71441-65186 Iran

Search for more papers by this author
R. Koebnik

R. Koebnik

IRD, CIRAD, Univ. Montpellier, Interactions Plantes Microorganismes Environnement (IPME), 34394 Montpellier, France

Search for more papers by this author
J. R. Lamichhane

J. R. Lamichhane

UMR AGIR, INRA, 31326 Castanet-Tolosan, France

Search for more papers by this author
First published: 31 October 2017
Citations: 18

Abstract

During 2013–2016, 277 tomato fields were surveyed across Iran to monitor the status of bacterial canker of tomato, caused by Clavibacter michiganensis subsp. michiganensis. Altogether, 450 plant samples were collected, both with and without symptoms, from which 35 bacterial strains were recovered. These were positive for the PCR test performed using the Clavibacter-specific primer pair CMR16F1/CMR16R1. Based on the phylogeny of the gyrB gene sequences, 31, three and one of the 35 strains were identified as C. michiganensis, Microbacterium sp. and Agrococcus sp., respectively. The 31 strains of C. michiganensis were further identified as C. michiganensis subsp. michiganensis (23 strains), C. michiganensis subsp. tessellarius (six strains) and Clavibacter spp. (two strains). This was subsequently confirmed by multilocus sequence analysis (MLSA) of five housekeeping genes (atpD, gyrB, ppk, recA and rpoB). In pathogenicity tests, all 23 strains induced wilting symptoms on tomato plants in greenhouse conditions, while no symptoms were observed on eggplant, bell pepper and chili pepper plants. All evaluated pathogenicity determinant genes (celA, pat-1, tomA, ppaA, chpC and chpG) were detected in 18 out of 31 C. michiganensis strains, using eight specific primer pairs. Estimation of the number of nucleotide differences, sequence similarity matrix and MLSA clustered two peach-coloured strains (Tom495 and Tom532) separately from all nine previously described subspecies, thereby suggesting these two strains are a new subspecies of C. michiganensis. However, a detailed taxonomic study using multiphased molecular approaches is needed to delineate a formal taxonomic name for these atypical strains.

Introduction

Bacterial canker of tomato (Solanum lycopersicum), caused by Clavibacter michiganensis subsp. michiganensis, is considered as an economically important disease affecting tomato production worldwide (Eichenlaub & Gartemann, 2011). The disease was first identified on tomato in Michigan (USA) in 1909 (Smith, 1910). Since then, the pathogen has been detected in several geographic regions, and to date, the disease has been reported from several continents around the globe (EPPO, 2016). The causal agent is included in the A2 list of quarantine pathogens by the European and Mediterranean Plant Protection Organization (EPPO, 2016). The pathogen is seedborne, and infected seeds are the main source of inoculum for long-distance dissemination and the consequent disease outbreaks (Jacques et al., 2012). The host range of C. michiganensis subsp. michiganensis comprises mainly solanaceous vegetables such as tomato, pepper (Capsicum sp.) and eggplant (Solanum melongena) (Eichenlaub & Gartemann, 2011).

Until a few years ago, all C. michiganensis strains associated with tomato and pepper were classified as C. michiganensis subsp. michiganensis (Eichenlaub & Gartemann, 2011; Yim et al., 2012). However, Jacques et al. (2012) conducted a series of molecular and physiological tests, including multilocus sequence typing (MLST), on nonpathogenic C. michiganensis strains isolated from tomato and pepper. The authors showed that these strains were distinct from C. michiganensis subsp. michiganensisbona fide’. Subsequently, comprehensive studies on the seed-associated C. michiganensis strains using biochemical tests, BIOLOG-based metabolic profiling, cell wall analyses and MLST led to the description of two new subspecies, namely C. michiganensis subsp. californiensis and C. michiganensis subsp. chilensis (Yasuhara-Bell & Alvarez, 2015). The latter study also highlighted that Clavibacter strains isolated from pepper have unique morphological features but are yet to be delineated as a new subspecies. Accordingly, C. michiganensis subsp. capsici has been recently described as the causal agent of bacterial canker of pepper (Oh et al., 2016). In addition to the three new subspecies of the pathogen associated with solanaceous vegetables, a new yellow-pigmented strain isolated from common bean (Phaseolus vulgaris) seeds and capable of causing bean leaf yellowing disease, has been designated as C. michiganensis subsp. phaseoli in Spain (Gonzalez & Trapiello, 2014).

The four subspecies of the pathogen associated with tomato and pepper, C. michiganensis subsp. michiganensis, C. michiganensis subsp. californiensis, C. michiganensis subsp. chilensis and C. michiganensis subsp. capsici, differ from each other in terms of genetic and phenotypic characteristics, including pathogenicity (Yasuhara-Bell & Alvarez, 2015; Oh et al., 2016). For instance, C. michiganensis subsp. capsici is more aggressive on pepper than on tomato, and produces orange-pigmented colonies with lower mucoidy on nutrient broth yeast extract (NBY) agar and yeast extract-dextrose-calcium carbonate (YDC) agar media (Yim et al., 2012). Furthermore, neither the plasmid-borne (i.e. pat-1 and celA) nor the pathogenicity island located (i.e. chpC, chpG, ppaA and tomA) virulence genes were detected in C. michiganensis subsp. capsici (Yim et al., 2012). Recently, the genome sequence based digital DNA–DNA hybridization and average nucleotide identity on five subspecies of C. michiganensis suggested a higher taxonomic position (species-level) for these bacteria (Tambong, 2017).

In Iran, bacterial canker of tomato was first reported in 1993 in West Azerbaijan province, in the northwestern part of the country, and since then the disease has occasionally been observed in the region (Mazarei et al., 1993; Nazari et al., 2007). At the time, the pathogen was only identified based on morphological and biochemical characteristics and was reported to have only yellow-pigmented colonies on NBY medium (Nazari et al., 2007). However, studies carried out during 2013–2016 throughout the country revealed the presence of a number of bacterial species associated with solanaceous vegetables (i.e. eggplant, pepper and tomato) which were divided into three groups: (i) Xanthomonas sp. strains, in particular X. euvesicatoria and X. perforans identified as the causal agents of leaf spot of pepper and tomato plants, respectively (Osdaghi et al., 2016, 2017a); (ii) epiphytic Curtobacterium flaccumfaciens strains isolated from symptomless solanaceous plants which were pathogenic on leguminous (e.g. common bean and cowpea) but not on solanaceous plants (Osdaghi et al., 2017b); and (iii) mucoidal Corynebacterium-like isolates resembling Clavibacter sp., which have not been characterized yet.

The objectives of the present study were therefore to characterize isolates from Iran using phenotypic and genetic approaches, including pathogenicity and detection of the pathogenicity determinant genes, as well as phylogenetic characterization using MLST/MLSA.

Materials and methods

Surveys, sampling and isolation

During the four consecutive years (2013–2016), several field surveys were conducted in Iran to monitor the distribution of bacterial diseases affecting solanaceous vegetables. Eggplant, pepper and tomato growing areas across central, northern, northwestern, southern, southwestern and western parts of the country were surveyed. The surveyed areas, number of fields and the number of samples collected in each area are described in Table 1. Solanaceous plants with suspected bacterial disease symptoms on leaves, stems, sepals and fruits were collected, and brought to the laboratory for further analysis.

Table 1. The number of fields surveyed in this study, their location and samples collected during each inspection. Clavibacter michiganensis subsp. michiganensis was isolated in East Azerbaijan, Qazvin, Kohgiluyeh-Boyer-Ahmad, West Azerbaijan and Zanjan provinces of Iran
Province County Number of Sampling date
Surveyed fields Samples Infected fields Isolates
East Azerbaijan Marand 5 7 0 0 August 2013
Khuzestan Shushtar 10 10 0 0 May 2014
Khuzestan Dezful 4 11 0 0 May 2014
Hormozgan Minab 3 10 0 0 January 2015
Bushehr Borazjan 18 23 0 0 March 2015
Bushehr Borazjana 4 11 0 0 March 2015
Fars Khesht 5 15 0 0 April 2015
Fars Kazerun 21 22 0 0 May 2015
Fars Shiraz 15 18 0 0 June 2015
East Azerbaijan Marand 15 17 4 7 August 2015
East Azerbaijan Azarshahr 5 16 0 0 August 2015
West Azerbaijan Urmia 10 18 3 4 August 2015
West Azerbaijan Mahabad 5 11 2 3 August 2015
Kurdistan Bijar 5 15 0 0 August 2015
Zanjan Zanjan 10 13 2 4 August 2015
Zanjan Abhar 3 9 0 0 August 2015
Qazvin Takestan 6 17 4 6 August 2015
Qazvin Qazvin 5 14 0 0 August 2015
Alborz Karaj 7 14 0 0 August 2015
Markazi Delijan 4 11 0 0 September 2015
Markazi Khomeyn 5 8 0 0 September 2015
Isfahan Golpayegan 3 7 0 0 September 2015
Fars Abadeh 3 9 0 0 September 2015
Fars Marvdasht 10 11 0 0 September 2015
Fars Persepolis 4 15 0 0 September 2015
Fars Bajgaha 2 4 0 0 November 2015
Fars Firuzabad 25 27 0 0 April 2016
Fars Firuzabad 15 16 0 0 May 2016
Fars Marvdasht 15 15 0 1 June 2016
Fars Kazerun 12 16 0 0 June 2016
Fars Kaftarak 4 6 0 0 June 2016
Kohgiluyeh-Boyer-Ahmad Yasuj 2 4 1 1 July 2016
Kohgiluyeh-Boyer-Ahmad Sisakht 3 5 1 2 July 2016
Kohgiluyeh-Boyer-Ahmad BoyerAhmad 4 8 0 3 July 2016
Golestan Kordkuy 5 9 0 1 August 2016
Golestan Gorgan 5 8 0 3 August 2016
  • a Samples were collected from tomato-producing greenhouses.

The samples were surface sterilized by dipping into 0.5% sodium hypochlorite for 20 s followed by two to three rinses in sterile distilled water (SDW). Small pieces of tissue were cut from the margin of the lesions with a sterile scalpel, and macerated in a few drops of SDW using a sterile mortar and pestle. A loopful of the resulting suspension was streaked onto yeast-extract peptone glucose agar (YPGA) and nutrient agar (NA) media as described by Schaad et al. (2001). The plates were incubated at 25–27 °C for 48–72 h. Pure cultures of the resulting bacterial isolates were obtained by colony subculturing. The obtained bacterial isolates were resuspended in SDW and stored at 4 °C for further use. For long-term storage, the isolates were maintained in 15% glycerol at 70 °C.

Phenotypic characterization of isolates

The purified bacterial isolates were divided into three groups based on their Gram stain reaction, colony colour and colony morphology (mucoidy or fluidity on YDC medium). Two groups of bacteria (Xanthomonas sp. and C. flaccumfaciens) were previously characterized and published elsewhere (Osdaghi et al., 2016, 2017a,2017b). The third group, which consisted of Gram-positive Clavibacter-like isolates, were subjected to biochemical and phenotypic tests including oxidase and catalase activity, aerobic/anaerobic growth (O/F), growth on 0.1% triphenyl tetrazolium chloride (TTC), as well as the hydrolysis of aesculin and Tween 80 (Schaad et al., 2001). The type strain of C. michiganensis subsp. michiganensis (ICMP 2550) was used as positive control. All biochemical tests were repeated twice.

Pathogenicity tests

All purified isolates (Table 2) were evaluated for their pathogenicity on four commonly grown solanaceous crops in Iran. Bell pepper ‘Sereno’, chili pepper ‘Aziz’, eggplant ‘Emami’, and tomato ‘Sunseed 6189’ plants were grown in 20 cm diameter pots (three plants per pot and three pots per species), and were maintained in the greenhouse at ambient conditions (24–27 °C, 14 h natural light). The plants were inoculated at 10–15 days post-emergence, when they had at least three fully expanded leaves (approximately 15 cm height). Inoculation was made by inserting a sterile dissecting needle dipped into a freshly prepared bacterial suspension on YDC medium (108 CFU mL−1 in SDW) through the first internode of each plant (EPPO, 2016). The inoculated plants were maintained in the greenhouse at ambient temperature (24–26 °C and 14 h natural light, 70–80% relative humidity). The plants were periodically monitored for the appearance of disease symptoms up to 30 days post-inoculation (dpi). The positive and negative control plants were treated in the same manner, using the type strain of C. michiganensis subsp. michiganensis (ICMP 2550) and SDW, respectively. Koch's postulates were accomplished by reisolating the inoculated strains on YPGA medium from all inoculated plants. Confirmation of the reisolated bacteria was made by determining Gram stain reaction and colony characteristics on YDC medium as well as by using the specific primer pair CMR16F1/CMR16R1 (Lee et al., 1997; Table S1). The pathogenicity tests were conducted twice.

Table 2. Bacterial strains used in this study, their origin, host and date of isolation, colony colour and pathogenicity on tomato plants. The representative strains from this study are deposited in the International Collection of Microorganisms from Plants (ICMP, Auckland, New Zealand)
Isolate ICMPa Identified as Host County Province Year Colony colour Pathogenicity on tomato
ICMP 2550cd ICMP 2550 Cmmb Tomato Hungary 1957 Yellow +
Tom110 Cmm Tomato Urmia West Azerbaijan 2015 Yellow +
Tom138 Cmm Tomato Takestan Qazvin 2015 Yellow +
Tom453 Cmm Tomato Urmia West Azerbaijan 2015 Yellow +
Tom457 Cmm Tomato Takestan Qazvin 2015 Yellow +
Tom460 Cmm Tomato Marand East Azerbaijan 2015 Yellow +
Tom463 Cmm Tomato Marand East Azerbaijan 2015 Yellow +
Tom464 Cmm Tomato Zanjan Zanjan 2015 Yellow +
Tom465 Cmm Tomato Zanjan Zanjan 2015 Yellow +
Tom466 Cmm Tomato Zanjan Zanjan 2015 Yellow +
Tom471 Cmm Tomato Takestan Qazvin 2015 Yellow +
Tom493 Agrococcus sp. Tomato Marand East Azerbaijan 2015 Yellow
Tom495d ICMP 22060 Clavibacter michiganensis Tomato Marand East Azerbaijan 2015 Peach
Tom498 Cmm Tomato Marand East Azerbaijan 2015 Yellow +
Tom506 Cmm Tomato Urmia West Azerbaijan 2015 Yellow +
Tom530 Cmm Tomato Takestan Qazvin 2015 Yellow +
Tom532d ICMP 22100 C. michiganensis Tomato Marand East Azerbaijan 2015 Peach
Tom666 Microbacterium sp. Tomato Marvdasht Fars 2016 Yellow
Tom808d ICMP 22050 Cmm Tomato Zanjan Zanjan 2015 Yellow +
Tom812 Cmm Tomato Urmia West Azerbaijan 2015 Yellow +
Tom823 Cmm Tomato Mahabad West Azerbaijan 2015 Yellow +
Tom826d ICMP 22051 Cmm Tomato Mahabad West Azerbaijan 2015 Yellow +
Tom835d ICMP 22052 Cmm Tomato Takestan Qazvin 2015 Yellow +
Tom946 Cmm Tomato Mahabad West Azerbaijan 2015 Yellow +
Tom947 Cmm Tomato Takestan Qazvin 2015 Yellow +
T210d Cmm Tomato Yasuj Kohgiluyeh-Boyer-Ahmad 2016 Yellow +
T220 C. michiganensis Tomato Sisakht Kohgiluyeh-Boyer-Ahmad 2016 Orange
T221 C. michiganensis Tomato Kordkuy Golestan 2016 Orange
T222 Cmm Tomato Sisakht Kohgiluyeh-Boyer-Ahmad 2016 Yellow +
T228 Microbacterium sp. Tomato BoyerAhmad Kohgiluyeh-Boyer-Ahmad 2016 Yellow
T229d C. michiganensis Tomato BoyerAhmad Kohgiluyeh-Boyer-Ahmad 2016 Orange
T236 C. michiganensis Tomato BoyerAhmad Kohgiluyeh-Boyer-Ahmad 2016 Orange
T31 Microbacterium sp. Tomato Gorgan Golestan 2016 Orange
TBd C. michiganensis Tomato Gorgan Golestan 2016 Orange
TC C. michiganensis Tomato Gorgan Golestan 2016 Orange
Zol2d ICMP 22049 Cmm Tomato Marand East Azerbaijan 2015 Yellow +
LPPA 982cd C. michiganensis subsp. phaseoli Common bean nd nd nd Yellow nd
  • nd, Not determined.
  • a International Collection of Microorganisms from Plants.
  • b Clavibacter michiganensis subsp. michiganensis.
  • c Type strain.
  • d Strains used for multilocus sequence typing.

Molecular characterization of the pathogen

Detection with specific PCRs

The bacterial isolates (Table 2) were tested using the primer pairs CMR16F1/CMR16R1 and PSA-4/PSA-R, specific for Clavibacter spp. and C. michiganensis subsp. michiganensis, respectively (Lee et al., 1997; Pastrik & Rainey, 1999; Table S1). DNA extraction was carried out using Expin Combo GP (GeneAll) DNA extraction kit, as recommended by the manufacturer. The quality and quantity of DNA were spectrophotometrically evaluated and concentration adjusted to 50 ng μL−1 using NanoDrop ND-100 (NanoDrop Technologies) for further use. For PCRs, Universal PCR kit Ampliqon Taq DNA Polymerase Master Mix Red (Ampliqon A/S) was applied according to the manufacturer's recommendations. For each strain, a 50 μL PCR, including 100 ng total DNA and 3 μL of each primer (10 pmol μL−1), was used. The sequences of the primer pairs and annealing temperatures are described in Table S1. In all molecular tests, the type strains of C. michiganensis subsp. michiganensis (ICMP 2550) and C. flaccumfaciens pv. flaccumfaciens (ICMP 2584) were used as positive and negative controls, respectively. The pure DNA of the type strain of the newly described subspecies C. michiganensis subsp. phaseoli (LMG 27667; Gonzalez & Trapiello, 2014; Table 2) was also used in all molecular and phylogenetic analyses.

Detection of pathogenicity determinant genes

The pathogenicity of C. michiganensis subsp. michiganensis depends on several pathogenicity determinant genes located on either pCM1/pCM2 plasmids (celA and pat-1) or the low G+C content pathogenicity island (tomA, ppaA, chpC and chpG) of the chromosomal DNA (Gartemann et al., 2008; Jacques et al., 2012; Yim et al., 2012). PCR tests were performed using eight primer pairs (Table S1) to evaluate the presence of the pathogenicity genes in the strains isolated in Iran. The presence of the plasmid-borne genes was determined using the cel-578up/cel-2752low, pCRcel-593/pCRcel-1860 and PFC3/PFC5 primer pairs specific for the celA gene, as well as the Cmm-5/Cmm-6 primer pair specific for the pat-1 gene (Dreier et al., 1995; Jahr et al., 2000; Kleitman et al., 2008). In addition, the primer pairs tomA-F/tomA-R, ppaA-F/ppaA-R, chpC-F/chpC-R and chpG-F/chpG-R were used for the detection of the tomA, ppaA, chpC and chpG genes, respectively, located on the pathogenicity island of the C. michiganensis subsp. michiganensis chromosomal DNA (Kleitman et al., 2008). All PCR conditions used were the same as described above. The sequences of the primer pairs and respective annealing temperatures are described in Table S1.

Phylogenetic analysis

The phylogeny of the gyrB gene has sufficient resolution and specificity to identify subspecies of C. michiganensis (Zaluga et al., 2011; Jacques et al., 2012). The bacterial isolates were thus subjected to phylogenetic analysis using the gyrB sequences. The primer pair 2F/6R, which amplifies 977 bp of the gyrB gene in the Microbacteriaceae family, was used in this study (Table S1; Richert et al., 2005). The PCR parameters were the same as described above. The certificated PCR products were sent to Bioneer Corporation (http//:www.Bioneer.com) to be sequenced via Sanger sequencing technology, and the resulting sequences were analysed with the blast program (http://blast.ncbi.nlm.nih.gov/). Sequences of the gyrB gene from a worldwide collection of C. michiganensis strains (Jacques et al., 2012; Yasuhara-Bell & Alvarez, 2015; Oh et al., 2016) were retrieved from the NCBI GenBank and included in the phylogenetic analysis. Phylogenetic trees were constructed using neighbour joining and maximum likelihood methods with mega v. 6.06 software (Tamura et al., 2013). The Jukes–Cantor distance model was used to generate a phylogenetic tree through the neighbour joining method, while the model of evolution for maximum likelihood analysis was determined using modeltest tab in mega v. 6.06 (Hall, 2011). Rathayibacter iranicus was used to root the phylogenetic trees, which were constructed with bootstrapping (1000 replications).

MLST/MLSA

To obtain precise and reliable data on the phylogenetic position of the bacterial isolates used in this study, 10 representative isolates (Table 2) were selected for MLST/MLSA. Five housekeeping genes (atpD, gyrB, ppk, recA and rpoB) were subjected to MLST analysis as recommended in the literature (Jacques et al., 2012; Yasuhara-Bell & Alvarez, 2015; Oh et al., 2016). The PCR conditions and sequencing procedure were the same as described above. The sequences of these genes for all C. michiganensis subspecies present in the international collections were obtained from the NCBI GenBank and included in the analysis (Gonzalez & Trapiello, 2014; Yasuhara-Bell & Alvarez, 2015; Oh et al., 2016). The sequences were concatenated following the alphabetic order of the genes, ending in a sequence of 2508 bp: nucleotides 1–518 for the atpD, 519–1072 for the gyrB (554 bp), 1073–1603 for the ppk (531 bp), 1604–2197 for the recA (594 bp) and 2198–2508 for the rpoB (311 bp) genes. Phylogenetic trees were constructed for all individual genes as well as the concatenated data set of sequences, as described above. Because of their atypical nature, the 16S rRNA gene of the two peach-coloured isolates (Tom532 and Tom495) was also sequenced using the fD1/rP2 primer pair (Weisburg et al., 1991) and phylogenetically compared with those of almost all other species of the Microbacteriaceae family (Evtushenko & Takeuchi, 2006).

The nucleotide diversity, number of haplotypes and minimum number of recombination events were determined using DnaSP v. 5.10 software (Librado & Rozas, 2009). The class I neutrality tests (Tajima's D, Fu and Li's D* and Fu and Li's F* statistics) were also performed to detect the departure from the mutation/drift equilibrium (Librado & Rozas, 2009). Splits-decomposition networks were constructed using SplitsTree v. 4.14.4 (Huson & Bryant, 2006). Split decomposition is a parsimony method that allows a tree-like network structure if conflicting phylogeny signals are detected in the data set. The similarity matrix of the concatenated sequences of five housekeeping genes in the type strains of the C. michiganensis subspecies was prepared using the online service ‘Sequence Identity And Similarity’ (SIAS, http://imed.med.ucm.es/Tools/sias.html) with default settings and BLOSUM62 matrix. Rathayibacter iranicus, the closest species to C. michiganensis, was used as an out-group. All obtained sequences were deposited in the NCBI GenBank and assigned accession numbers.

Results

In total, 277 fields were surveyed across the country and 450 plant samples (either with or without symptoms) were collected (Table 1). While 72% of the surveyed fields were cultivated only to tomato, the remaining 28% contained eggplant, pepper and tomato plants (mixed cropping). Wilting symptoms were observed on tomato plants in East Azerbaijan, Kohgiluyeh-Boyer-Ahmad, Qazvin, West Azerbaijan and Zanjan provinces. Dull green symptoms with oily areas in the interveinal regions of the leaves were observed which desiccated over time. Marginal pale brown necrotic areas had appeared on the leaves, with a scorched appearance. Symptoms on tomato fruits were rarely observed and were seen only in the West Azerbaijan province, as small and slightly raised lesions with a white halo. No whole plant wilting, defoliation or destructive disease outbreaks and extensive yield losses were observed in the surveyed fields. In total, 17 out of 277 fields showed bacterial wilt symptoms and 51 Gram-positive Corynebacterium-like bacterial isolates were recovered from the samples collected across the surveyed areas.

Of 51 Corynebacterium-like bacterial isolates, 16 were previously characterized and published as C. flaccumfaciens (Osdaghi et al., 2017b), and thus were excluded from this study. The remaining 35 isolates resembled Clavibacter-like bacteria with mucoidal colonies on YDC medium. Of these 35 isolates, 26 were yellow-pigmented, seven were orange-pigmented, and two isolates produced peach-coloured colonies on YDC medium (Table 2). All isolates were negative for oxidase, anaerobic growth and hydrolysis of Tween 80, while they were positive for catalase, hydrolysis of aesculin and growth on TTC medium.

Pathogenicity tests

Of 35 isolates tested, 23 induced wilting symptoms on tomato (Table 2) and all of these had yellow-pigmented colonies. The inoculated tomato plants showed leaf wilting 10–12 dpi and almost died 20–25 dpi. The orange- and peach-coloured isolates did not cause disease symptoms on the inoculated plants. None of the isolates induced disease symptoms on eggplant, bell pepper and chili pepper. The inoculated bacterial isolates were consistently reisolated from plants with symptoms that were inoculated, and their identity was confirmed using both morphological and biochemical characteristics, as well as the specific primer pair CMR16F1/CMR16R1 as described above (data not shown). The control plants inoculated with SDW remained symptomless.

Molecular characterization of the isolates

The primer pair CMR16F1/CMR16R1 directed the amplification of a 1425 bp DNA fragment in all 35 bacterial isolates (Table 3). However, the C. michiganensis subsp. michiganensis-specific primer pair PSA-4/PSA-R amplified the expected 271 bp DNA fragment only in 28 out of 35 isolates. The results were negative for isolates Tom110, Tom493, Tom495, Tom532, T228, T31, and the amplified fragment in isolate Tom666 was longer than the expected size (≥350 bp; data not shown). As for the detection of pathogenicity determinant genes, all tested primer pairs directed the amplification of the expected DNA fragments in 18 out of 35 isolates (Table 3). However, in some cases the expected DNA fragment was not amplified, for instance in strains Tom463 (with tomA-F/tomA-R primer pair), Tom823 (with chpG-F/chpG-R primer pair), Tom946 (with tomA-F/tomA-R and chpG-F/chpG-R primer pairs), as well as T210 (with cel-578up/cel-2752low primer pair; Table 3). In contrast, none of the primer pairs amplified the expected DNA fragment in isolates T220, T221, T222, T228, T229, T236, T31, TB and TC.

Table 3. Results of the PCR tests for the detection of eight pathogenicity determinant genes in Clavibacter michiganensis strains obtained in this study. The two right-hand columns indicate the results of the PCR tests using genus- and subspecies-specific primer pairs
Strain Colony colour Pathogenicity on tomato cel-578up/cel-2752lowa pCRcel-593/pCRcel-1860b PFC3/PFC5c tomA-F/tomA-Rd ppaA-F/ppaA-Re chpC-F/chpC-Rf chpG-F/chpG-Rg Cmm-5/Cmm-6h PSA-4/PSA-Ri CMR16F1/CMR16R1j
ICMP 2550k Yellow + + + + + + + + + + +
Tom110 Yellow + + + + + + + + + +
Tom138 Yellow + + + + + + + + + + +
Tom453 Yellow + + + + + + + + + + +
Tom457 Yellow + + + + + + + + + + +
Tom460 Yellow + + l + + + + + + + + +
Tom463 Yellow + + l + + + + + + + +
Tom464 Yellow + + + + + + + + + + +
Tom465 Yellow + + + + + + + + + + +
Tom466 Yellow + + + + + + + + + + +
Tom471 Yellow + + + + + + + + + + +
Tom493 Yellow + + l +
Tom495 Peach + + + l + l +
Tom498 Yellow + + + + + + + + + + +
Tom506 Yellow + + + + + + + + + + +
Tom530 Yellow + + + + + + + + + + +
Tom532 Peach + + + l + l + l +
Tom666 Yellow + m +
Tom808 Yellow + + l + + + + + + + + +
Tom812 Yellow + + + + + + + + + + +
Tom823 Yellow + + + + + l + + + + +
Tom826 Yellow + + + + + + + + + + +
Tom835 Yellow + + + + + + + + + + +
Tom946 Yellow + + + + + + + + +
Tom947 Yellow + + + + + + + + + + +
T210 Yellow + + + + + + + + + +
T220 Orange + +
T221 Orange + +
T222 Yellow + + +
T228 Yellow +
T229 Orange + +
T236 Orange + +
T31 Orange +
TB Orange + +
TC Orange + +
Zol2 Yellow + + l + + + + + + + + +
LPPA 982k Yellow nd + l + l + + l + +
  • nd, Not determined.
  • Primer pairs for the deletion of: acelA, bcatalytic domain of celA, ccellulose-binding domain of celA, dtomA, eppaA, fchpC, gchpG, hpat-1, iC. michiganensis subsp. michiganensis, and jClavibacter spp.
  • kType strain.
  • lFaint amplification.
  • mThe length of the DNA fragment was longer than the expected amplicon.

Phylogenetic analysis

blast search using the gyrB gene sequences on the NCBI GenBank database confirmed 31 out of the 35 isolates studied as C. michiganensis (99–100% sequence similarity). Isolate Tom493 was identified as Agrococcus sp. (with 99% gyrB sequence similarity; accession number MF409300), while isolates Tom666 (MF409320 and MF409344 for the ppk and rpoB genes, respectively), T228 (MF409319, MF409331 and MF409343 for the ppk, recA and rpoB genes, respectively), and T31 (MF409275 and MF409342 for the atpD and rpoB genes, respectively) were identified as Microbacterium sp. (data not shown). Because of the negative results in pathogenicity tests, and their irrelevant identity, isolates Tom493, Tom666, T228 and T31 were excluded from the phylogenetic analysis.

The gyrB gene phylogeny, supported with high bootstrap values, confirmed all 23 isolates as C. michiganensis subsp. michiganensis, all of which produced yellow-pigmented mucoidal colonies on YDC medium (Fig. 1; Table 2). While the orange-pigmented isolates TC, T220, TB, T236, T221 and T229 clustered within the members of C. michiganensis subsp. tessellarius (Fig. 1), the type strain of C. michiganensis subsp. phaseoli (LMG 27667) clustered within the strains of C. michiganensis subsp. chilensis. Interestingly, two isolates (Tom495 and Tom532), both producing peach-coloured colonies, clustered separately from all subspecies of C. michiganensis (Fig. 1).

Details are in the caption following the image
Phylogeny of Clavibacter michiganensis strains obtained in this study using gyrB gene sequences. Maximum likelihood method based on the general time reversible (GTR) model was used. Percentage bootstrap values >50% from 1000 samplings are indicated. Rathayibacter iranicus was used as an out-group cluster. The numbers in the parentheses indicate the number of clonal complex sequences/strains in each branch. Of the 31 C. michiganensis strains, 23 clustered within C. michiganensis subsp. michiganensis and six clustered within C. michiganensis subsp. tessellarius. Two peach-coloured strains (Tom495 and Tom532) clustered separately from all nine previously described subspecies of C. michiganensis. The strains isolated in Iran are labelled with a black triangle.

Phylogenetic trees constructed using the data set of concatenated sequences of the five housekeeping genes confirmed the results obtained from the phylogeny of the gyrB gene (Figs S1a,b, 2). All five strains (Zol2, Tom808, Tom835, T210 and Tom826) which identified as C. michiganensis subsp. michiganensis based on the gyrB gene phylogeny, clustered within C. michiganensis subsp. michiganensis strains, although they were scattered through different multilocus haplotypes of the subspecies (Fig. 2). Strains T229 and TB clustered within C. michiganensis subsp. tessellarius strains. Both the neighbour joining and maximum likelihood methods used for the MLST analysis provided similar phylogeny (Fig. S1a, b). Surprisingly, two peach-coloured strains Tom495 and Tom532 clustered separately and formed a unique group, differentiated from nine previously defined subspecies of C. michiganensis. Similar results were obtained when the sequences of the individual housekeeping genes were subjected to phylogenetic analysis (Fig. S1c–f). To further evaluate the phylogenetic position of atypical strains Tom495 and Tom532, and to decipher whether these strains still belonged to the genus Clavibacter, a 1256 bp fragment of 16S rRNA gene sequences was subjected to blast search. The latter showed these strains having 100% sequence similarity with the type strain of C. michiganensis subsp. capsici (PF008). A phylogenetic tree based on the 16S rRNA gene sequences of almost all species/subspecies of Microbacteriaceae confirmed the inclusion of strains Tom495 and Tom532 within C. michiganensis (Fig. S2).

Details are in the caption following the image
Phylogenetic tree based on the concatenated sequences of atpD, gyrB, ppk, recA and rpoB genes of Clavibacter michiganensis strains obtained in this study. Maximum likelihood method was used for the construction of the tree using general time reversible (GTR) model. Bootstrap scores (1000 replicates) are displayed at each node. Rathayibacter iranicus was used for rooting the tree. The numbers in the parentheses indicate the number of clonal complex sequences/strains in each branch. While the strains Zol2, Tom808, Tom835, T210 and Tom826 were clustered within C. michiganensis subsp. michiganensis subspecies, the strains T229 and TB clustered within C. michiganensis subsp. tessellarius. The strains Tom532 and Tom495 clustered separately and formed a unique group distinct from the nine previously defined C. michiganensis subspecies. The strains isolated in Iran are labelled with a black triangle.

The nucleotide differences between strain Tom532 and each of the type strains of nine subspecies were similar to, or even higher than, the differences observed among the nine subspecies (Table S2a–e). For instance, there were 24, 20, 21, 23, 27, 18, 20, 23 and 16 nucleotide differences in the gyrB gene sequences between strain Tom532 and the type strains of C. michiganensis subsp. michiganensis, C. michiganensis subsp. californiensis, C. michiganensis subsp. sepedonicus, C. michiganensis subsp. nebraskensis, C. michiganensis subsp. insidiosus, C. michiganensis subsp. chilensis, C. michiganensis subsp. phaseoli, C. michiganensis subsp. tessellarius and C. michiganensis subsp. capsici, respectively (Table S2b). Similar results were observed in the recA gene sequences (Table S2d). Only one, four, zero, one and three nucleotide differences were observed between the type strains of C. michiganensis subsp. chilensis and C. michiganensis subsp. phaseoli in the atpD, gyrB, ppk, recA, and rpoB gene sequences, respectively (Table S2a–e).

Overall, 70 multilocus haplotypes were observed among 122 C. michiganensis strains used for the phylogenetic analysis (Table 4). Five multilocus haplotypes were observed among C. michiganensis subsp. michiganensis strains isolated in Iran (Zol2, Tom808, Tom835, T210 and Tom826), indicating that these strains are not monophyletic. Fu and Li's D* statistics (1.23407*, < 0.05) showed that there was significant departure from the mutation drift equilibrium within C. michiganensis subsp. michiganensis strains isolated in Iran (Table 4). The nucleotide diversity analysis showed a greater sequence variation among C. michiganensis subsp. tessellarius strains used in this study (ND = 0.01868; Table 4) than the variability observed among other strains from other subspecies.

Table 4. Sequence variation statistics and diversity parameters for the Clavibacter michiganensis strains used in this study
Sequence set Strains Haplotypes Total no. segregating sites Nucleotide diversity (π) θwa No. mutations (η) Haplotype (gene) diversity Neutrality tests R b
Tajima's D Fu and Li's D* Fu and Li's F*
All 122 70 313 0.02008 0.02321 360 0.978 0.82403, NS 0.80240, NS 0.97293, NS 67
C. michiganensis subsp. michiganensis 74 32 58 0.00427 0.00474 61 0.945 0.48137, NS 2.09309, NS 1.74817, NS 6
C. michiganensis subsp. michiganensis (Iran) 5 4 21 0.00335 0.00402 21 0.900 1.23407, NS 1.23407*, < 0.05 1.32325, NS 0
C. michiganensis subsp. californiensis 6 6 44 0.00625 0.00768 44 1.000 1.19531, NS 1.19196, NS 1.30679, NS 3
C. michiganensis. subsp. sepedonicus 4 4 8 0.00179 0.00174 8 1.000 0.30903, NS 0.30903, NS 0.30417, NS 0
C. michiganensis subsp. nebraskensis 5 3 5 0.00088 0.00096 5 0.700 0.56199, NS 0.56199, NS 0.57792, NS 0
C. michiganensis subsp. insidiosus 4 2 3 0.00060 0.00065 3 0.500 0.75445, NS 0.75445, NS 0.67466, NS 0
C. michiganensis subsp. chilensis 8 8 45 0.00600 0.00692 45 1.000 0.71730, NS 0.84088, NS 0.90553, NS 4
C. michiganensis subsp. tessellarius 6 5 107 0.01584 0.01868 108 0.933 1.03452, NS 1.02905, NS 1.13217, NS 4
C. michiganensis subsp. capsici 6 5 38 0.00659 0.00664 38 0.933 0.04174, NS 0.08413, NS 0.06268, NS 2
Tom532 and Tom495 2 1 0 0.00000 0.00000 0 0.000 nd nd nd nd
  • NS, not significant; nd, not determined.
  • a Watterson's theta.
  • b Minimum number of recombination events.

Because the maximum likelihood phylogenies showed incompatible topologies (Figs 1, 2 & S1), phylogenetic networks were generated using the splits-decomposition method for all individual genes (data not shown), as well as the data set of the concatenated sequences, which confirmed a recombination among the strains (Fig. 3). Strains Tom495 and Tom532 clustered separately from all previously described subspecies, confirming their atypical nature (Fig. 3). The orange-pigmented strain TB, which clustered within C. michiganensis subsp. tessellarius strains in the maximum likelihood trees, clustered separately from all members of this subspecies in the splits-decomposition network (Fig. 3). Additionally, the strain C74A, which belonged to C. michiganensis subsp. californiensis subspecies in the maximum likelihood phylogenetic tree, clustered separately in the splits-decomposition network (Fig. 3). The similarity matrix of the concatenated sequence of the five housekeeping gene sequences showed strain Tom532 having 96.61, 96.81, 96.53, 96.57, 96.53, 97.00, 97.20, 96.33 and 97.04% sequence similarity with the type strains of C. michiganensis subsp. michiganensis, C. michiganensis subsp. californiensis, C. michiganensis subsp. sepedonicus, C. michiganensis subsp. nebraskensis, C. michiganensis subsp. insidiosus, C. michiganensis subsp. phaseoli, C. michiganensis subsp. chilensis, C. michiganensis subsp. tessellarius and C. michiganensis subsp. capsici, respectively (Table 5). Interestingly, the type strains of C. michiganensis subsp. phaseoli and C. michiganensis subsp. chilensis showed 99.64% sequence similarity in the five housekeeping gene sequences (Table 5).

Details are in the caption following the image
Splits-decomposition network generated from the concatenated sequences of atpD, gyrB, ppk, recA and rpoB genes of Clavibacter michiganensis strains obtained in this study. The parallel lines indicate conflicting phylogenetic relationships and recombination among the strains. The strains Tom532 and Tom495 clustered separately from the nine previously defined subspecies of C. michiganensis.
Table 5. Similarity matrix of the concatenated sequences of five housekeeping genes (atpD, gyrB, ppk, recA and rpoB) in the type strains of all Clavibacter michiganensis subspecies. Rathayibacter iranicus was used as the out-group
Similarity matrix (%) Tom532 Cmm Cmcali Cms Cmn Cmi Cmp Cmchi Cmt Cmcap R. iran.
Tom532 100
C. michiganensis subsp. michiganensis (Cmm) 96.63 100
C. michiganensis subsp. californiensis (Cmcali) 96.82 98.21 100
C. michiganensis subsp. sepedonicus (Cms) 96.55 96.94 97.06 100
C. michiganensis subsp. nebraskensis (Cmn) 96.59 96.59 96.78 96.35 100
C. michiganensis subsp. insidiosus (Cmi) 96.55 96.59 96.67 96.98 98.09 100
C. michiganensis subsp phaseoli (Cmp) 97.02 96.82 96.98 96.51 97.74 97.42 100
C. michiganensis subsp. chilensis (Cmchi) 97.22 97.02 97.1 96.71 97.93 97.62 99.64 100
C. michiganensis subsp. tessellarius (Cmt) 96.35 96.31 96.47 96.35 96.39 96.47 96.31 96.51 100
C. michiganensis subsp. capsici (Cmcap) 97.06 97.02 97.22 96.55 96.74 96.86 96.9 97.02 96.67 100
Rathayibacter iranicus (R. iran.) 82.91 82.16 82.16 81.92 82.16 81.76 82.28 82.4 82.12 82.24 100

The sequenced nucleotides were deposited in the GenBank database under the following accession numbers: atpD: MF409265–MF409275, gyrB: MF409276–MF409308, ppk: MF409309–MF409320, recA: MF409321–MF409331, rpoB: MF409332–MF409344, and 16S rRNA: MF409345–MF409346. A pure culture of the representative strains (i.e. Tom495, Tom532, Tom808, Tom826, Tom835 and Zol2) is deposited in the International Collection of Microorganisms from Plants (ICMP, Landcare Research, Auckland, New Zealand) and assigned accession numbers (Table 2).

Discussion

The results of this study showed that bacterial canker of tomato is present across central and southern provinces of Iran. Previously the disease has been reported only from the west Azerbaijan and Golestan provinces (Nazari et al., 2007) as the survey was not conducted across other provinces of the country. During the present surveys, no disease outbreaks leading to severe economic losses were observed throughout the study areas. However, given the seedborne nature of the pathogen (Eichenlaub & Gartemann, 2011), quarantine restrictions should be applied to prevent its further distribution towards major tomato producing areas of the country, and, in particular, in Fars and Bushehr provinces, southern Iran (Osdaghi et al., 2017a).

The Clavibacter-specific primer pair CMR16F1/CMR16R1, derived from the 16S rRNA gene sequences of C. michiganensis strains (Lee et al., 1997), directed the amplification of a 1425 bp DNA fragment in Corynebacterium-like bacterial strains including Agrococcus sp. and Microbacterium sp. (Table 3). Because the primer pair CMR16F1/CMR16R1 has not been assessed yet for its specificity towards a large number of Microbacteriaceae strains, it could not be used for the specific detection of Clavibacter sp. strains as recommended by Yim et al. (2012). False negative and false positive results were also obtained when the C. michiganensis subsp. michiganensis-specific primer pair PSA-4/PSA-R was used (Table 3). The latter primer pair failed to detect strain Tom110, identified as C. michiganensis subsp. michiganensis based on the gyrB gene sequence. Furthermore, this primer pair amplified the 271 bp DNA fragment in the orange-pigmented and nonpathogenic strains T220, T221, T229, TC and TB (Table 3). This result emphasizes that the primer pair PSA-4/PSA-R should not be used as a sole detection method for the screening of plant materials (i.e. seed samples) or bacterial strains in quarantine inspections. Indeed, previous studies recommended a set of diagnostic methods, including specific primer pairs, pathogenicity tests and phylogenetic analysis using the gyrB gene sequences, for a precise and accurate identification of Clavibacter sp. strains associated with solanaceous vegetables (Jacques et al., 2012; EPPO, 2016).

The plasmid-borne pathogenicity determinant genes (celA and pat-1) were detected in all pathogenic strains. An exception was strain T210 in which the celA gene was not detected, despite the presence of the catalytic and cellulose-binding domains of the celA gene. A weak amplification of the celA, and the pat-1 genes were observed in the atypical strains Tom495 and Tom532, indicating that these strains might carry the pCM1/pCM2 plasmids (Table 3; Gartemann et al., 2008). Because no catalytic and cellulose-binding domains of the celA and the tomA gene were detected in strains Tom495 and Tom532, it can be hypothesized that the partial deletion of the plasmid-borne genes and the chromosomal pathogenicity island are behind the avirulent nature of these strains. However, only further in-depth studies on the genetic content of these strains may provide insights in this regard.

Among the diseases caused by the subspecies of C. michiganensis, only bacterial canker of tomato and bacterial wilt of alfalfa, caused by C. michiganensis subsp. michiganensis and subsp. insidiosus, respectively, were officially reported in Iran (Nazari et al., 2007; Heidari & Khodakaramian, 2011). The present study identified the orange-pigmented strains TC, T220, TB, T236, T221 and T229 as C. michiganensis subsp. tessellarius using both the individual gyrB gene sequences as well as the MLST analysis. Although C. michiganensis subsp. tessellarius did not cause disease symptoms on plant species tested in this study, this bacterium is the causal agent of leaf freckles and leaf spots of wheat (Eichenlaub & Gartemann, 2011). Therefore, further studies may be useful to test the pathogenicity of these strains on wheat. In addition, specific field surveys would be useful to know the presence of this pathogen across the wheat growing areas of the country.

The most interesting results obtained in this study are that the atypical peach-coloured strains Tom495 and Tom532 clustered separately from all nine C. michiganensis subspecies, based on the MLST analysis. At the same time, the phylogenetic analysis based on the 16S rRNA genes sequences confirmed the inclusion of these strains within C. michiganensis species. The results of the phylogenetic analysis and similarity matrix, using the sequences of five housekeeping genes, suggest that these strains could be considered as a new subspecies of C. michiganensis. The sequence similarity between strain Tom532 and other nine subspecies varied from 96.35% to 97.22% for C. michiganensis subsp. tessellarius and C. michiganensis subsp. chilensis, respectively. Notably, the type strains of C. michiganensis subsp. chilensis and C. michiganensis subsp. phaseoli showed 99.64% sequence similarity while the type strains of C. michiganensis subsp. michiganensis and C. michiganensis subsp. californiensis had 98.21% sequence similarity. This further supports the possible assignment of strains Tom495 and Tom532 as a new subspecies. However, additional biochemical tests, DNA–DNA hybridization, BIOLOG-based metabolic profiling and cell wall analyses are needed for the description of a new subspecies (Yasuhara-Bell & Alvarez, 2015).

During the last 3 years, four new subspecies of C. michiganensis have been described, including C. michiganensis subsp. phaseoli, the causal agent of bean leaf yellowing disease (Gonzalez & Trapiello, 2014), C. michiganensis subsp. californiensis and C. michiganensis subsp. chilensis as seed-associated nonpathogenic bacteria (Yasuhara-Bell & Alvarez, 2015), and C. michiganensis subsp. capsici causing bacterial canker in pepper (Oh et al., 2016). Results of the present study showed that the type strains of C. michiganensis subsp. phaseoli (LMG 27667) and C. michiganensis subsp. chilensis (CFBP 8217) are phylogenetically close to each other and cannot be distinguished using the MLST/MLSA approach (Fig. 2). A misunderstanding seems to have occurred in the description of C. michiganensis subsp. chilensis, because the type strain of C. michiganensis subsp. phaseoli has not been included in the phenotypic and phylogenetic studies by Yasuhara-Bell & Alvarez (2015). Recently, using the digital DNA–DNA hybridization on the genome sequences of five C. michiganensis subspecies, Tambong (2017) proposed a reconsideration in the taxonomy of the species. Furthermore, the average nucleotide identity (ANI) analysis using all available completely sequenced C. michiganensis subspecies showed that there are great variations in the ANI values among C. michiganensisbona fide’. For instance, the ANI between the nonpathogenic C. michiganensis strain CASJ009 isolated from tomato in California, USA (Thapa et al., 2017) and C. michiganensis subsp. michiganensis, C. michiganensis subsp. insidiosus, C. michiganensis subsp. nebraskensis, C. michiganensis subsp. sepedonicus, C. michiganensis subsp. capsici and C. michiganensis subsp. tessellarius, is only 88%. The ANI of C. michiganensis subsp. capsici and C. michiganensis subsp. tessellarius with all above-mentioned subspecies is 90% (E. Osdaghi, unpublished data). These ANI values are far below the accepted threshold (95–96%) for the definition of prokaryotic species (Kim et al., 2014). A comprehensive multiphased taxonomic study using the type strains of all existing C. michiganensis subspecies, as well as strains Tom532 (isolated in this study) and CASJ009 (isolated in California; Thapa et al., 2017), is warranted to decipher the precise taxonomic status of the species.

Based on the MLST analysis, Jacques et al. (2012) reported that C. michiganensis subsp. michiganensis strains isolated in different countries or even different continents over the years are monophyletic and are distinct from all other subspecies of C. michiganensis. The present results confirmed that C. michiganensis subsp. michiganensis is a distinct and monophyletic clade; however, there are significant genetic variations within the subspecies. For instance, five representative C. michiganensis subsp. michiganensis strains isolated in Iran were clustered in five distinct multilocus haplotypes based on the MLSA data. This indicates that the bacterial canker pathogen has been introduced to Iran from different sources. Until recently, there has been no attempt to monitor the presence of the bacterial canker pathogen in Iran, except for the restricted study conducted by Nazari et al. (2007), and so the history of the pathogen in Iran and the date of entry into the country is unclear. Because C. michiganensis subsp. michiganensis strains isolated in Iran clustered with the strains isolated in Algeria, Belgium, Brazil and Slovenia, it can be assumed that the bacterial canker pathogen in Iran has been introduced from these areas, or vice versa, thereby emphasizing the importance of the seedborne nature of the pathogen and its propagation worldwide.

Acknowledgments

The authors thank Ana Gonzalez (SERIDA, Villaviciosa, Asturias, Spain) for providing the pure DNA of the type strain of C. michiganensis subsp. phaseoli (LMG 27667). This study was financially supported by Shiraz University.