Introduction

Tomato (Solanum lycopersicum L., Family: Solanaceae) is one of the most consumed vegetable crops cultivated worldwide. Tomato is an excellent source of human nutrition and health, because of the richness of vitamins, minerals, and fibers (Kumar et al. 2022). The major tomato-growing countries are China, India, Turkey, The United States of America, Egypt, Italy, and Iran (FAOSTAT 2022). Tomato is cultivated in most regions of Iran, because of its favorable climate; especially in the northwest and south where it is produced traditionally. Several factors, such as insects, plant pathogens, weeds, climatic conditions, and unsuitable soils can reduce the yield of different tomato cultivars (Siddique et al. 2020). Up to now, 67 plant pathogenic agents including 15 bacteria that are known to cause diseases in tomato. Bacterial diseases are one of the main reasons for the decreasing quantity and quality of tomato yield in the world (Jones et al. 2014). Clavibacter michiganensis subsp. michiganensis (Cmm), Xanthomonas vesicatoria, X. euvesicatoria, X. perforans, X. gardneri, Pseudomonas syringae pv. tomato and Ralstonia solanacearum are the common tomato pathogens reported globally and can cause serious damages. Control of mentioned diseases is difficult and requires integrated disease management strategies to reduce the damage (Blancard 2012). Cmm is one of the most dangerous and economically important plant pathogens in the world, it is present the list of the quarantine plant pathogens of the European Union and many other countries in Asia, Africa, and the Caribbean (Blancard 2012; Sen et al. 2015).

For the first time in 1909, the disease was reported from a greenhouse in Michigan, US; and was described in 1910 by Smith. As the disease is seed-borne, it spread rapidly to other states and epidemics occurred in the north and west of the US, and in some states of Canada in the decades of 1930, 1960, and 1980 (Borkar and Yumlembam 2017; Sen et al. 2015). Following the epidemics, research was conducted on various aspects of the disease. In Iran, according to existing reports tomato bacterial canker was reported from different parts of the country, and Cmm has a high potential for damage to this product (Nazari et al. 2007; Osdaghi et al. 2018).

Infected seed is the primary source of infection. In secondary dispersion, Cmm enters plant tissue through the stomata, hydathodes, roots, and damaged tissues, and it reaches the xylem vessels resulting in systemic infections (Valenzuela et al. 2018). Pathogenicity is mediated by virulence factors and transcriptional regulators encoded by the pathogenicity island (PAI) on a chromosome and two natural plasmids (pCM1 and pCM2) (Nandi et al. 2018). Cmm produces several types of degrading enzymes as virulence factors, such as serine proteases, cellulases, pectinases, and xylanases. These enzymes have essential roles in pathogenicity (Hwang et al. 2019). Cell wall hydrolyzing enzymes, especially cellulose, provide nutrient materials for the pathogenic bacterium (Bella et al. 2012; Valenzuela et al. 2021). Symptoms begin mainly as unilateral wilt of the leaves that later spread to entire the foliage. Cmm produces tiny, raised, and white blisters on the surface of young leaves and fruits. On the fruit’s surface, the centers of the white blisters turn brown giving rise to “bird’s eye” spots, and necrotic lesions at the edges of older leaves are usually visible. In a wilted plant, the color of the vascular tissues near the nodes changes to dark yellow or tan brown. Cankers develop on the stems and infected plants die in advanced stages (Bella et al. 2012; Siddique et al. 2020), although symptomless latent infections and the invasion of tomato seeds by Cmm are widespread (Nandi et al. 2018).

A yield reduction of up to 84% has been reported due to the tomato plants contaminated by Cmm. In vitro condition, the rate of disease damage varied from 46 to 93% (Sen et al. 2015). The prevalence of the disease symptoms in several greenhouses in Italy led to identify Cmm as the disease agent (Bella et al. 2012). Identification of Cmm was done by studying phenotypic traits and using primers PSA4/PSAR and PSA8/PSAR (Pastrik and Rainey 1999). In addition, the genetic diversity of the isolates was studied in rep-PCR using BOX, ERIC, and REP primers (Wassermann et al. 2017). Analysis of fatty acids via GC-FAME method and PCR using specific primers of CMM5/CMM6 were successfully used to identify Cmm isolates in the Mediterranean regions of Turkey (Basim and Basim 2018).

Preventing and management of seed-borne and soil-borne diseases such as tomato bacterial canker is very difficult. In addition, no available resistant cultivar has been introduced for this disease. Therefore, disease management should be done with a combination of preventive and control methods. Usage of certified seed, sterilizing equipment and crop rotation for at least 1–3 years, removing symptomatic seedlings and plants, deeply plowing infected derbies, spraying of copper-base compounds and antibiotics such as streptomycin are mentioned for the integrated management of the disease (Christopher Peritore-Galve et al. 2021; de Leon et al. 2011).

The purpose of this study was identifying the disease agent, which is the first and most crucial step in designing management strategies for the control or reducing the bacterial disease in tomato cultivation. For this, the identification of pathogenic isolates was performed based on phenotypic and molecular characteristics; and the phylogenetic relationship of the isolates was determined using 16 S rRNA gene sequence.

Materials and methods

Sampling and bacterial isolation

During the summer of 2016–2018, sampling of symptomatic tomato plants was done in infected fields in West Azarbaijan province, northwest of Iran. Disease symptoms include leaf spots and necrotic areas on leaves, discolored and brown lines in xylem vascular tissues, and stem canker. The symptomatic plants were placed in paper bags. During the transfer, the samples were placed in a chilled container, and were stored in a refrigerator at 4 °C. Bacteria were isolated from symptomatic leaves and stems. The infected leaves and stems were surface-disinfected by using 70% ethanol for 30 s and 2% sodium hypochlorite for 5 min and then washed thrice with sterile distilled water (SDW). Young spots on leaves were picked up with a disinfected scalpel blade, cut into 1–2 mm sections, and transferred to SDW. In the case of stems with canker symptoms, small sections of discolored vascular tissues were separated and transferred into SDW. After 15–20 min, serial dilutions were prepared and a loopful of the suspension 10−3 CFU mL−1 was cultured on nutrient broth yeast agar (NBYA) medium and Petri dishes were incubated at 28 °C for a week. A total of 387 bacterial isolations were done from infected samples in 48–96 h after incubation. Future experiments on isolates were carried out with thrice purified isolates by streak plate method on yeast dextrose carbonate agar (YDC) medium. Purified bacterial cultures were obtained with homogeneity colony morphology. For long-term storage, bacterial cultures were placed in nutrient broth with 25% glycerol and were stored at − 80 °C. For routine experiments, bacterial isolates were cultured on nutrient agar (NA) medium or in Luria-Bertani broth (LB) medium at 28 °C for two days (Schaad et al. 2001).

Phenotypic experiments

Morphological, biochemical, and physiological features of studied bacteria were performed via standard bacteriological procedures. Tobacco (Nicotiana tabacum L.) and four o’clock flower (Mirabilis jalapa L.) were inoculated to record hypersensitivity reaction (HR) according to Ftayeh et al. (2010), Gram staining, colony morphology study on nutrient-broth yeast extract agar (NBY), and yeast dextrose carbonate agar (YDC) media were done. Further, the growth on triphenyl tetrazolium chloride agar (TTC) and on Clavibacter nebraskense selective (CNS) media was done. Levan, oxidase, motility, arginine dihydrolase, tyrosinase activity and hydrolysis of casein, esculin, gelatin, tween 80 and starch were carried out according to Schaad et al. (2001) and Klement et al. (1990). Aerobic/anaerobic growth, growth at 4 and 40 °C, production of H2S from peptone, tolerance to 4, 5, and 7% NaCl (WV-1), indole production, nitrate reduction, phosphatase, urease, lecithinase, and catalase were performed according to Schaad et al. (2001). The strain of C. michiganensis subsp. michiganensis NCPPB 382, provided by the National Collection of Plant Pathogenic Bacteria (UK), was used as a reference strain.

Pathogenicity and virulence tests

The infection procedure was performed following de Ftayeh et al. (2010). Tomato seedlings (cv. Super Stone) were grown in 20 cm diameter pots in a soil mix (containing sand, perlite, and peat compost in equal proportion) under natural daylight in the greenhouse. Six-week-old tomato seedlings were pierced on the main stem between the cotyledons and the first true leaf with a sterile scalpel that was dipped in a fresh bacterial colony. For virulence tests, 35 µL of the bacterial suspension in SDW of 109 CFU mL−1 was injected into the axil of the second or third true leaf. After inoculation, tomato plants were covered with clear polyethylene bags for 24 h at room temperature (18 °C). After removing the bags, infected plants were maintained in a greenhouse with a controlled climate at 28 °C under natural daylight with 80% relative humidity (RH). NCPPB 382 and SDW were used as positive and negative controls, respectively. Infected plants were monitored for vigor and developed symptoms up to 21 days after inoculation.

To fulfill Koch’s postulates, the pathogen was re-isolated from the inoculated plants showing disease symptoms and re-identified via phenotypic tests including Gram staining, aerobic growth, colony morphology, and pigment production on YDC, as well as PCR using specific primers of CMM5/CMM6.

The strain’s disease severity was assessed on the inoculated plants according to symptoms on a scoring of 1–5 (1—no disease symptom, 2—wilting in 1–25% of the plant, 3—wilting in 26–50% of the plant, 4—wilting in 51–75% of the plant, and 5—wilting in 76–100% of plant or plant death). Four replicates were considered for each strain (Klement et al. 1990). The experiment was repeated twice.

The disease severity (DS) for each strain was calculated using the following formula (Bella et al. 2012):

$$\text{D}\text{S}=\frac{\sum \left(\text{n}\times {\text{x}}_{1-5}\right)}{\text{N}}\times 100,$$

where DS (%) is the percentage of disease severity, n is the number of plants in each numerical score, x1 − 5 is the numerical score, and N is the total number of evaluated plants for each strain.

Antibiogram

An antibiogram test was performed using sterile paper disks impregnated with antibiotics. Ten pathogenic Cmm strains from different regions and with different features in phenotypic, pathogenic, and virulence characters were selected for this test. The susceptibility reaction of selected strains was evaluated against 22 different antibiotics prepared from Padtan Teb Laboratory Instruments (Iran) (Table 1). 100 µL of bacterial suspension (1 × 109 CFU mL−1) was spread over NA medium supplemented with 1% (WV−1) glucose in 10 cm Petri dishes (Azucena et al. 2019). Then, antibiotic disks were placed 25 cm from each other on the medium and incubated at 28 °C for 48 h. A blank disk with SWD was used as the negative control. The inhibition zone was measured 48 h after incubation. The experiment was conducted with a completely randomized design (CRD). Four repeats were considered for each antibiotic disk. Statistical analysis of data was performed by Tukey test, at the 99% level of confidence (p < 0.01), using SAS software (version 9.4). The experiment was repeated thrice.

Table 1 Concentration of antibiotic in prepared disks used in antibiogram test
Full size table

Molecular identification

DNA extraction

Bacterial genomic DNA was extracted based on Li and de Boer (1995). A suspension of active bacterial colonies in 200 µL SDW was prepared. Then, 400 µL extraction buffer (Tris-HCl 50mM, pH: 8; EDTA 25mM; SDS 1%; Proteinase K 10 µg mL−1) was added, and tubes were placed in Ban Mari at 55 °C for 3 h. Then, 400 µL ammonium acetate (7.5 M) was added to tubes, and DNA was recovered by isopropanol. DNA was re-dissolved in SDW after washing the plates with ethanol 70% twice. The quantity of DNA was adjusted to 50 ng µL−1. The extracted DNA was stored at − 20 °C.

PCR using universal and specific primers

Molecular identification was done using a pair of universal primers fD1/rP2 and fD1/rD1 (Weisburg et al. 1991) and three sets of specific primers including PSA4/PSAR (Pastrik and Rainey 1999), PSA8/PSAR (PSA8 is a modification of PSA4), and CMM5/CMM6 (Dreier et al. 1995) (Table 2). FD1/rP2 and fD1/rD1 were derived from 16 S rRNA gene, PSA4/PSAR, and PSA4/PSAR were designed based on the 16-23 S rRNA intergenic spacer region, and CMM5/CMM6 was derived from a part of the pat-1 gene localized on plasmid pCM2 (70.0 Kbp).

PCR reactions were performed with a 25 µL PCR cocktail containing 12 µL of Taq DNA Polymerase 2x Master Mix RED (Amplicon, Denmark), 1 µL of each primer (10 pmol µL−1), and 1 µL of template DNA (50 ng). PCR amplification program for fD1/rP2 or fD1/rD1 primers was carried out under the following conditions; initial denaturation cycle at 94 °C (5 min), 35 cycles of denaturation at 94 °C (1 min), annealing at 62 °C (1 min) and extension at 72 °C (1.5 min), and a final extension at 72 °C for 7 min. PCR condition for specific primers was initial denaturation cycle at 94 °C (5 min), 30 cycles of denaturation at 94 °C (50 s), annealing at 58 °C (PSA4/PSAR and PSA4/PSA8) and 60 °C (CMM5/CMM6) (1 min) and extension at 72 °C (50 s), and a final extension at 72 °C for 7 min in a Palm cycler model GP001 (Corbett Research Co., Australia. For all samples, six µL of PCR product was applied in 1.0% agarose gel at 75 V for one hour, and Gene Ruler TM 1 kb DNA ladder was used as a marker. SDW was used as a negative control. Agarose gel 1.0% was stained by FluoroDye DNA Fluorescent Loading Dye 1 µL mL−1 (Smobio, Taiwan). Gels were photographed by a gel documentation system model ENDURO™ GDS (Industrial Labnet, USA).

Table 2 Traits of primers used in this study
Full size table

Phylogenetic tree using 16 S rRNA sequence

Amplified DNA fragments were purified with an Expin™ Combo GP kit (GeneAll, South Korea), and sequenced using Sanger sequencing technology Macrogen Corporation (https://www.macrogenusa.com) (Seoul, South Korea). The sequences were edited using the Chromas software. To perform multiple sequence alignment, the GenBank database of nucleic acid sequences, which showed high similarity to extracted 16 S rRNA, was used. Alignment and comparison of the sequences were performed with ClustalX programs available on the bioinformatics web databases. The NCBI BLAST (the Basic Alignment Search Tool) database was used to analyze sequence homology for extracted 16 S rRNA gene sequences. Phylogenetic relations were inferred, by applying the Kimura-2-parameter model. The neighbor-joining (NJ) method and the adjacent method by MEGA 11 software (Tamura et al. 2021) were used for the phylogenetic tree and tested by bootstrap (1000 repetitions).

Results

Symptoms and phenotypic characteristics

Disease symptoms such as dark brown to black lesions on the edges of leaves, xylem discoloration and plant wilting, cankers on stems and petioles, and death of plants were observed in infected tomato plants (Fig. 1).

Fig. 1
figure 1

Symptoms of bacterial canker disease on tomato plant. Necrosis on edge of leaves (a), plant wilting (b), stem canker and brown lines on xylem tissues of stem (c), leaves necrosis and stem canker (d), and dead infected plant by Clavibacter michiganensis pv. michganensis (e)

Full size image

A total of 56 bacterial isolates out of 387 were found gram-positive, obligate aerobic, and non-motile. The colonies were semifluid and their color was yellow on YDC but different colors in a range of yellowish-orange colors were observed on the NBY medium. Hypersensitive reaction on tobacco and four o’clock flower leaves was positive with the formation of necrotic spots at the injection site of the isolates, but production of levan polymer, oxidase, arginine dihydrolase, nitrate reduction, and urease were negative. All isolates grew on TTC and CNS media, as well as NA containing NaCl 4–7%. Hydrolysis of starch, tween 80, and casein was negative but hydrolysis of aesculin was positive; and the reaction of isolates in gelatin liquefaction was variable. Other phenotypic results including carbon source utilization are listed in Table 3.

Table 3 Phenotypic features of pathogenic C. michiganensis subsp. michiganensis isolated from tomato plants
Full size table

Pathogenicity and virulence

Results of the pathogenicity test on tomato seedlings indicated that 35 strains out of 56 Cmm strains were virulent with different degrees of disease symptoms on tomato seedlings. Other strains with positive HR on tobacco and four o’clock flower leaves did not show any disease symptoms on tomato plants. Among the virulent Cmm, the disease severity of 14 strains was more than 55%, so these strains were categorized as virulent strains. Disease severity from 21 to 55% was observed in 13 strains with moderated virulence. Other eight strains showed poor virulence with disease severity of less than 20% and in some cases, mild symptoms were noticed after two weeks (Table 4). Phenotypic features and PCR results of the re-isolated bacteria from inoculated plants proved Koch’s postulates.

Table 4 Place and year of isolation, and severity of disease caused by pathogenic Clavibacter michiganensis pv. michganensis used in this study
Full size table

Molecular identification

To confirm the phenotypic identification, PCR was conducted using two sets of universal primers fD1/rP2 and fD1/rD1, and three sets of specific primers of PSA4/PSAR, PSA8/PSAR, along with CMM5/CMM6.

A 1500 bp fragment was amplified in all pathogenic and non-pathogenic Cmm strains (Fig. 2).

Fig. 2
figure 2

The visualization of PCR amplification product from 16 S rRNA gene of Clavibacter michiganensis subsp. michiganensis strains. L) 1.0 Kb DNA ladder, (1) NCPPB382, (2) cmm.Ur.42, (3) cmm.Mh24, (4) cmm.Mi38, (5) cmm.Ur93, (6) cmm.Mh13, (7) cmm.Kh5, (8) cmm.Ur91, and C: negative control

Full size image

The retrieved results of 16 S rRNA sequences of the studied strains revealed 99.2–100% similarity to C. michiganensis subsp. michiganensis registered in GenBank (NCBI). Sequences of eight Cmm strains were deposited in GenBank databases under the accession numbers shown in Table 5.

Table 5 Accession numbers of 16 S rRNA sequences of selected Clavibacter michiganensis pv. michganensis in GenBank (NCBI)
Full size table

A 270 bp fragment was amplified using PSA4/PSAR in 33 isolates out of 56 isolates which presented a typical Cmm colony; however, PSA8/PSAR primers identified 47 isolates by amplification of a fragment with a similar size (Fig. 3). On PCR using CMM5/CMM6 primers, 24 isolates were identified and the amplified fragment was 614 bp (Fig. 4) (Table 4).

Fig. 3
figure 3

Polymerase chain reaction analysis of ITS between 16 S rRNA and 23 S rRNA in pathogenic (1) NCPPB382, 2) cmm.Ur.16, (4) cmm.Mi88, 6) cmm.Kh55, (7) cmm.Mh13, (8) cmm.Mi.17) and non-pathogenic, (3) cmm.Mi80, (5) cmm.Ur.80, (9) cmm.Kh5) Clavibacter michiganensis subsp. michiganensis strains. L: 1.0 Kb DNA ladder, and C: negative control

Full size image
Fig. 4
figure 4

Polymerase chain reaction analysis of pat-1 gene in pathogenic Clavibacter michiganensis subsp. michiganensis strains. L) 1.0 Kb DNA ladder, (1) NCPPB382, (2) cmm.Kh53, (3) cmm.Mi17, (4) cmm.Ur9, (5) cmm.Mi88, and C) negative control

Full size image

A BLAST search in the GenBank database (NCBI) indicated that the sequences 16 S rRNA gene from the eight strains of this research had 99.2–100% homology with those of reference strains of C. michiganensis subsp. michiganensis. A Neighbor-joining phylogenetic tree of 16 S rRNA partial sequences from selected Cmm strains showed that the tomato isolates were grouped in two clads with 20 reference strains from the database. In clad A, the strains of cmm.Mi17, cmm.Ur42, cmm.Ur221, cmm.Ur147, cmm.Kh53, cmm.Ur9, and cmm.Mi82 were grouped with Cmm strains deposited from Germany, France, Italy, South Korea, Russia, China, Egypt, Japan, Kenya, Hungry, USA, and United Kingdom; in the clad B, cmm.Mh8 and four strains from South Korea and China were grouped. Both clads were separated from Clavibacter michiganensis pv. nebrankensis ATCC 27,822 which was used as an outgroup strain (Fig. 5).

Based on the results of phenotypic, pathogenic, and molecular experiments using universal and specific primers, and in comparison with bacterial characters described in valid bacteriological references, the pathogenic strain was subsequently identified as Clavibacter michiganensis subsp. michiganensis.

Fig. 5
figure 5

Neighbor-joining phylogenetic tree of 16 S rRNA partial sequences from selected Clavibacter michiganensis pv. michiganensis strains. Sequences of the compared strains were obtained from GenBank. Clavibacter michiganensis pv. nebrankensis ATCC 27822 used as an out group strain. Bootstrap values was calculated from 1000 replicates

Full size image

Antibiogram

In the antibiogram experiment, ten pathogenic Cmm strains showed different susceptibility to 22 studied antibiotics. Based on the results of variance analysis, a statistically significant difference at 1% level was observed among the studied antibiotics (Table 6).

Table 6 Analysis of variance of antibiogram test
Full size table

Nalidixic acid, gentamicin, doxycycline, streptomycin, erythromycin, cloxacillin, rifampin, cephalexin, and furazolidone had the highest inhibition effect on bacterial growth compared to the control, respectively. Studied strains were resistant to other 13 antibiotics and no inhibition zone was formed around antibiotic discs (Fig. 6).

Fig. 6
figure 6

Inhibition growth zone around antibiotic disks. Data are the average of ten Clavibacter michiganensis pv. michganensis strains in four replicates. Different letters represent significant differences at p < 0.01

Full size image

Discussion

Recently, typical symptoms of bacterial canker disease were observed in tomato fields in northwest of Iran. The pathogenic bacterial isolates obtained from symptomatic plants were relatively uniform and showed slight differences in phenotypic properties. Biochemical and physiological characteristics indicated that the strains belong to the subspecies of C. michiganensis subsp. michiganensis.

Several phenotypic characters such as colony pigment production on YDC, aesculin hydrolysis, and production of H2S from peptone were similar to Cmm strains reported before; however, differences were observed in the results of gelatin liquefaction, starch hydrolysis, and phosphatase test (Nazari et al. 2007). The consumption of carbon sources including lactose, maltose, mannitol, raffinose, dolisitol, and sorbitol by the strains was different from those of previous studies (Nazari et al. 2007). These results suggest that the differences in the physiological and biochemical characteristics among the strains can be due to the adaptation of pathogens to plant hosts; furthermore, various strains obtained from different regions indicate the genetic diversity of populations.

Chemical control of Cmm is limited to a few copper-based formulations and antibiotics. The frequent application of these materials faces the emergence of resistant strains. Resistance of plant pathogenic bacteria to antibiotics was reported in previous studies (de Leon et al. 2008; Sundin and Wang 2018; Valenzuela et al. 2019). Our results determine that there was a sensitivity of pathogenic Cmm strains to 22 antibiotics. The results indicated that the strains are resistant to 13 antibiotics and these results can be useful in suggesting management strategies.

The pathogenic Cmm strains showed different degrees of disease severity on tomato seedlings. About 40% of pathogenic strains were determined as virulence strains, while other strains were moderate or poorly virulent. The pathogen-reducing effect can be due to the loss of genes involved in pathogenicity. More than ten genes such as chpC, pat-1, phpA, and phpB (encode serine-proteases), CAZymes (encode carbohydrate-active enzymes), celA and celB (encode cellulases), xysA and xysB (encode xylanases), pelA1 and pelA2 (encode pectinases), and tomA (encode tomatinase) are involved in the virulence of Cmm strains (Hwang et al. 2019; Nandi et al. 2018; Valenzuela et al. 2021).

Three sets of primer pairs were used for molecular identification of 56 gram-positive and HR-positive strains. Based on the target fragment amplification, the efficiency of PSA4/PSAR and PSA8/PSAR were 59% and 84%, respectively. PSA4/PSAR and PSA8/PSAR were designed according to a region between the 16–23 S rRNA intergenic spacer and 23 S rRNA gene, respectively; thus, these primers can identify pathogenic and non-pathogenic Cmm strains. The results of the current study are in agreement with those of Pastrik and Rainey (1999) that the modified primer pairs of PSA8/PSAR have more efficient in the detection of Cmm strains. CMM5/CMM6 was designed according to pat-1 gene (Burokiene 2006) and allowed the identification of the pathogenic Cmm strains. This gene is localized on plasmid pCM2 and encodes an enzyme of 280 amino acids with a molecular mass of 29.7 kDa. Serine-protease is an essential enzyme for the virulence of Cmm strains (Dreier et al. 1995). In the present study, pat-1 was detected in 42% of Cmm-like colonies strains, but this fragment did not amplify in 12 strains with positive pathogenicity test. According to the results of this study and previous findings (Bella et al. 2012), pathogenicity is controlled by several genes and related factors, as in some undetected strains with CMM5/CMM6, severe symptoms were observed in the pathogenicity test. Lastly, Valenzuela et al. (2021) revealed correlations between the degree of disease severity and the presence/absence of genes associated with virulence in pathogenic Cmm strains. Gene sequence analysis from the genomes of highly virulent strains indicated the presence of ten genes involved in virulence; while four of these genes, including celB, xysA, pat-1, and phpA, were not amplified in the strains causing mild symptoms. Yim et al. (2012) reported in PCR according to pat-1 and celA, two plasmid-borne virulence genes, no fragment was amplified in Cmm strains isolated from the peppers. In addition, pathogenicity-related genes located on a pathogenicity island (PAI) including chpC, chpG, ppaA, or tomA genes were not detected in pathogenic Cmm strains isolated from pepper but these genes were detected in tomato strains.

The phylogenic analysis based on 16 S rRNA gene of studied strains revealed high homology (99.2–100%) to C. michiganensis subsp. michiganensis registered in GenBank (NCBI). Neighbor-joining phylogenetic tree of 16 S rRNA partial sequences indicated that the studied Cmm strains are similar to those strains from different countries including Germany, France, Italy, South Korea, Russia, China, Egypt, Japan, Kenia, Hungry, USA, and the United Kingdom. This finding emphasizes the importance of the disease transmission by infected seeds and its spread in different regions.

Based on the results of this study, the pathogenic agent of tomato disease in North West of Iran is C. michiganensis subsp. michiganensis. Considering the seed-borne nature of Cmm strains, quarantine restrictions should be applied to prevent its distribution to other tomato-producing areas of the country. In addition, the pathogenic strains showed resistance to some antibiotics. Information about the sensitivity characteristics of the pathogenic strains isolated in a geographical region against chemical compounds or antibiotics can benefit population diversity studies and provide low-risk strategies in plant disease management.