Xanthomonas campestris is a gram negative rod-shaped bacterium, which infects a wide range of host plants. It has been classified into three pathovars (Pv.) namely campestris, raphani and incanae based on host specificity (Fargier and Manceau 2007; Roux et al. 2015). The race structures of Xanthomonas campestris pv. campestric (Xcc) and Xanthomonas campestris pv. raphani (Xcr) were confirmed through gene-for-gene model by determining relationship between host plant and pathogen. So far, 11 Xcc and three Xcr pathogenic races have been identified (Fargier and Manceau 2007; Vicente and Holub 2013). Till now, race structure of Xci remains unreported (Vicente et al. 2002; Fargier and Manceau 2007).

Bacterial blight is a major disease of ornamental crops, such as stock (Matthiola incana), wallflower (Erysimum cheiri) and candytuft (Iberis umbellata) (Pirone 1978; Vicente et al. 2001) of which are economically important in Holland, Italy, Japan, China, Australia and Israel (Ecker et al. 1995). Xci causes bacterial blight is strictly host-specific compared to the pathovars of Xcc and Xcr. Xci is either a seed- or soil-borne pathogen, which penetrates through wound or leaf hydathodes, and invades the vascular system of the flowering plant (Kendeick and Baker 1942; Williams 1980). The disease is readily get transmitted among healthy plants through water splash. The leaves show yellowing and wilting upon transmission (Fargier and Manceau 2007; Cruz et al. 2017). Initially symptoms arise on the main stem and lateral branches of leaf and gradually reaches leaf petioles and seed peduncles (Rahimian and Okhovatian 1989). The disease may cause 90% of total yield loss in ornamental crops (Minardi et al. 1988). Therefore, disease management is crucial to check the crop loss. Early diagnosis of the disease and its causal agent is vital for effective disease management. Although molecular techniques, like genomic fingerprinting (repetitive PCR) helps to distinguish X. campestris isolates at pathovar level (Vicente et al. 2006; Parkinson et al. 2009). Such technique requires several repetitions to obtain accuriate results. Recently, Multi-Locus Sequence Typing (MLST) has been used to determine the evolutionary relationship of different Xcc populations (Fargier et al. 2011; Cruz et al. 2018; Laala et al. 2021). But, MLST requires systemic steps such as data collection, data analysis and finally multi-loci sequence analysis. It is expensive, time consuming, need skilled personnel and sometimes unspeculated results could make ambiguity in distinguishing the bacterial strains (Zhang et al. 2004). Therefore, there is a need to adopt a new or an improved older technique for the robust detection of Xci in host plants.

Comparative genomic approach represents direct comparison of the collective genomes of a particular group of organisms either across the genus or species level. It reveals the similarities, differences and evolutionary patterns among the diverse organisms at genomic level (Abby and Daubin 2007; Sivashankari and Shanmughavel 2007). This approach was successful in understanding the evolutionary mechanism of Bacillus licheniformis, the comparison of homologous genes among Drosophila melanogaster and Homo sapiens (Rey et al. 2004). The term ‘pangenomics’ was introduced in the study of collective gene content of an organisms from the same genus or species (Garrigues et al. 2013). Pangenome analysis focusses on the functional distribution of genes into core genome (genes are common in all individuals), dispensable genome (genes are absent in at least one individual) or singletons (genes are found in only one individual) (Tettelin et al. 2005; Mira et al. 2010). Pangenome is significant in an evolutionary context (Reno et al. 2009) but it could also be used in a broader genomic aspect (Reinhardt et al. 2009). The importance of pangenome analysis is accomplished herein to identify genes or regions for developing Xci-specific marker.

To date, no molecular marker has been reported to detect Xci strain. PCR (polymerase chain reaction) techniques are rapid, sensitive, reliable and economical in the specific identification of the pathogens (Berg et al. 2005; Kałużna et al. 2014). Utilizing PCR-based molecular marker might be a precise method in the detection of Xci as like the previously reported Xcc (Rubel et al. 2019b), Xcc-race specific (Rubel et al. 2017; Afrin et al. 2018) and Xcr specific markers (Rubel et al. 2019a). In the Xci, type III effector (T3E) proteins are very essential for it’s virulence (Roux et al. 2015) to the host plants, which has been focused in this study for marker development. So far, the genomes of two Xci strains (CFBP 1606R and CFBP 2527R) are available and used in present study. The aim of this study is to develop an Xci-specific molecular marker to recognize them from other bacterial species. In this regard, the complete genomes of Xci (CFBP1606R and CFBP2527R) was compared across Xcc (B100, ATCC33913 CFBP5817, 8004 and CN17), Xcr (756C) and other Xanthomonas genomes using EDGAR 3.0 (Electronic Data Gathering, Analysis, and Retrieval) (https://edgar3.computational.bio.uni-giessen.de/cgi-bin/edgar_login.cgi) to harvest singleton genes from Xci genome for developing pathovar-specific markers. Xci-specific markers could be facilitated to the farmers in identifying the Xci strain within a short time for taking appropriate control measures against bacterial blight disease in ornamental crucifers.

Culture of bacterial strains

A total of 18 diverse bacterial strains was collected from different sources and used in this study (Supplementary Table S1). Bacterial strains were grown on King’s B media at 30°C for 48 hours. All bacterial strains were stored at –80°C in a liquid medium with 50% glycerol for further uses.

Extraction of bacterial DNA

Bacterial DNA was extracted from cultured bacterial cells using DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instruction. The quality of isolated DNA was checked using 1.2% agarose gel electrophoresis at 100 V and quantified using a NanoDropND-1000 spectrophotometer (NanoDrop, Wilmington, DE, USA).

Comparative genome analysis of Xcc, Xcr and Xci

Comparative genome analysis was done by using EDGAR 3.0 (Blom et al. 2009; Blom et al. 2016; Dieckmann et al. 2021) to understand the similarities and variations among the bacterial species. The genomes of Xcc (B100, ATCC33913, CFBP5817, CN14, CN15, CN16, 8009), Xcr (756C), Xci (CFBP2527R, CFBP1606R), X. axonopodis pv. campestris, X. euvesicatoria, X. axonopodis pv. diffenbachiae (LMG695), X. citri pv. glycines (CFBP2526) were collected from NCBI (National Center for Biotechnology Information) and used for comparative genome analysis (Table 1).

Table 1 . Xanthomonas chromosomes description including genome size, total gene and protein of each individual strain with NCBI taxonomic ID.

Strain	Xcc B100	Xcc ATCC33913	Xcc CFBP5817	Xcc 8004	Xcc CN17	Xcr 756C	Xci CFBP1606R	Xci CFBP2527R	Xev LMG930	Xcg CFBP2526	Xad LMG695	Xac Xac 29
Size (Mbp)	5.07	5.07	4.97	5.14	4.99	4.94	4.96	4.92	5.07	5.363	4.99	4.60
Gene	4,542	4,240	4,447	4,584	4,398	4,372	4,423	4,380	4,977	4,856	4,592	4,641
Protein	4,236	4,179	4,122	4,297	4,086	4,055	4,071	4,032	4,505	4,461	3,919	4,237
NCBI ID	NC_010688	NC_003902	NZ_CM002673	NC_007086	CP011946.1	NC_017271	NZ_CM002635	NZ_CM002636	NZ_CP018467	NZ_CM002268	NZ_CP014347	NC_020800

Xcc: X. campestris pv. campestris, Xcr: X. campestris pv. raphani, Xci: X. campestris pv. incanae, Xev: X. euvesicatoria, Xcg: X. citri pv. glycines, Xad: X. axonopodis pv. diffenbachiae and Xa: X. axonopodis [Source: https://www.ncbi. nlm.nih.gov/genome/genomes/151].

The EDGAR tool produced information on core genome, dispensable genome and singletons; which were considered for marker development. In addition, genome plot and the gene cluster (Fig. 1) representation were also obtained from the EDGAR tool.

Figure 1. Circular plot representation of gene clusters present in the X. campestris and X. citri pathovars.

Designing of primers for PCR validation

Xci specific singleton genes with a T3E were selected for designing primers. Primers were designed from ten randomly selected singletons genes using primer3plus (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi). The uniqueness of Xci-specific genes was confirmed by using the Xcc-BLAST tool (http://210.110.86.160/lab/home.html) and BLASTn of NCBI (www.ncbi.nlm.nih.gov). Total elven primers were designed, among them one primer pair was designed from each of the gene except XCICFBP2527_RS21720, from which two primers were designed due to large gene size. The prefix XCI_1F/R, XCI_2F/R, XCI_3F/R, XCI_4F/R, XCI_5F/R, XCI_6F/R, XCI_7F/R, XCI_8F/R, XCI_9F/R, XCI_10F/ R and XCI_11F/R were denoted for the 11 primers designed from randomly selected 10 singleton genes of Xci strain (Table 2).

Table 2 . List of primer sequences with expected product size and annealing temperatures for specific amplification of Xanthomonas campestris pv. incanae.

Primers	Forward (5’-3’)	Reverse (5’-3’)	Expected size (bp)	Annealing temperature (°C)
XCI_1F/R	ATGTCGATCCAGCTCAGGTG	TCAGGTTGCAGGAGGGCTTGG	495	64
XCI_2F/R	CCTTCAGGTCATGCACCATTG	ATCCGCATTCTGCCCTCCCCCCT	503	65
XCI_3F/R	GCGATGGATACAGTGACTGGAAG	CAACTGCAAGGCCGTATCGTA	612	62
XCI_4F/R	CAGGCGGCCGATGCGAAGG	TGATGTTGAAGAGTGTGCGGAC	604	63
XCI_5F/R	CATTCTCGCGGAGCGGCGTCTGC	AGCTGCTGAGTTGCCGCGCTCT	665	68
XCI_6F/R	TCTGGCTCCCACTATGTGC	CGGGATGGATGAGGCTGTCC	468	60
XCI_7F/R	GTTCGGACTGGCGCTCTGGC	TCTTCATCATCACTAACCCA	303	60
XCI_8F/R	GATCATGGTGTCGCCCATG	TTAAGCCACTTTGAGCTGCACC	324	60
XCI_9F/R	GTGACCCTGCGATACATCGG	GGTGCTGCCAAGGCCGGAGC	769	65
XCI_10F/R	GGTACTACTGGATGCCAACC	TCAGGCCTTCATAAGATTTTG	397	57
XCI_11F/R	ATGGATGCCGCTGCCGGTCG	GCGGGTAGGTGAGGCCGTCGA	595	65

PCR validation for specific detection of Xci

Specificity of the designed primer was determined by the PCR assay using extracted DNA from strain of each of the Xci and Xcr, eight strains from Xcc, four strains from Xanthomonas and four other test bacteria (Supplementary Table S1). PCR was carried out with 20 µL reaction mixtures contained 1.0 µL of forward and 1.0 µL reverse primers (10 pmol), 8.0 µL Emerald PCR master mix (Takara, Shiga, Japan), 9.0 µL ultra-pure water and 1.0 µL DNA (70 ng). The PCR condition was fixed as; the initial denaturation at 95°C for 5 minutes followed by 25 cycles of denaturation at 95°C for 30 seconds with annealing temperatures varied from 57°C to 68°C for 40 seconds and extension at 72°C for 45 seconds and final extension at 72°C for 5 minutes. The amplified PCR products were separated by electrophoresis using 8 µL of PCR products in 1.4% agarose gel dissolved in 1x TBE buffer and stained with HIQ Blue Mango (20,000X) (bioD, Seoul, South Korea) and run for 40 minutes with 100 V before visualizing under UV light (302 nm).

Sensitivity of the PCR

The bacterial DNA sample was diluted sequentially in ten fold increments using sterile double distilled water up to a dilution of 10^-9 and fixed the DNA concentration ranged from 7 ng/µL to 0.00000007 ng/µL for detection of the sensitivity level of the well amplified Xci specific primers. From each of the diluted samples, 1 µL DNA was used as a source of template DNA with the similar PCR mixture as well as conditions were described previously.

Xci specific genes from genome comparison

Comparative genomic approach was implied to identify the genetic similarities and variations among the diverse Xanthomonas subspecies and their strains. The overall description of the bacterial strains; such as genome size (bp), genes per genome/strain, protein per genome/strain and NCBI taxonomic ID are enlisted in the Table 1. The complete genomes of closely related strains of Xcc, Xcr and Xci were analysed by comparative genomics and a total of 4,024 genes were identifed in the Xci strain

(Fig. 2). Comparative genome analysis conferred the segregation of the genomes into three major categories: 1) core genome, 2) dispensable genome and 3) singletons. The core genome consisted of 70% of the total gene (2,834) which are conserved across the genome of the bacterial strains, dispensable or flexible genome contained 1,108 genes which covered about 28% of the total number of genes and the remaining 2% comprised 82 singleton genes, which were found only in Xci genome (Fig. 2). These 82 singleton genes were retrieved from EDGAR tool and confirmed the Xci specificity by Xcc-BLAST tool and BLASTn of NCBI. Out of 82 singleton genes, only 10 genes were selected randomly for Xci-specific primer designing. Among the selected genes, only Xci specific four genes encoded hypothetical proteins and one gene encoded T3E protein (Table 3).

Table 3 . Molecular marker information for specific detection of Xanthomonas campestris pv. incanae.

Primers	Gene name	Start	End	Length (bp)	Direction	Protein description
XCI_1F/R	XCICFBP2527_RS16430	3641561	3641067	495	-	Hypothetical protein
XCI_2F/R	XCICFBP2527_RS13555	2955263	2956369	1,107	+	Hypothetical protein
XCI_3F/R	XCICFBP2527_RS21720	4861522	4862766	1,245	+	Hypothetical protein
XCI_5F/R	XCICFBP2527_RS03935	886319	884361	1,959	-	Hypothetical protein
XCI_6F/R	XCICFBP2527_RS21980	196812	196171	642	-	Type III effector

Figure 2. Classification of core (2834), dispensable (1108) and singleton (82) genes. Pie charts showing the distribution of genes in Xanthomonas campestris pv. incanae genome in three segments: core genome (70%) (blue); dispensable genome (28%) (red) and singletons (2%) genes (green).

Specificity test for detecting Xci strain

Out of eleven primers, five primer pairs; XCI_1F/R, XCI_2F/R, XCI_3F/R, XCI_5F/R and XCI_6F/R were amplified with expected PCR amplicons of 495 bp, 503 bp, 612 bp, 665 bp and 468 bp, respectively in the Xci population

(Fig. 3, Supplementary Table S1). Those amplifications were completely absent in other Xanthomonas spp. such as X. campestris pv. campestris, X. campestris pv. raphani, X. euvesicatoria, X. axonopodis pv. glycines, X. campestris pv. zinniae, X. campestris or in P. syringae pv. maculicola, Erwinia carotovora subsp. carotovora, Plasmodiophora brassicae and Didymella bryoniae. These results confirmed that 5 primer sets are specific to Xci only. The primer sets, XCI_1F/R, XCI_2F/R, XCI_3F/R and XCI_5F/R covered the gene loci of hypothetical proteins; XCICFBP2527_RS16430, XCICFBP2527_RS13555, XCICFBP2527_RS21720 and XCICFBP2527_RS03935, respectively. Primer set XCI_6F/R tagged T3E protein encoding locus XCICFBP2527_RS21980. The remaining 6 primers neither produced target amplicons nor even amplified and concluded as redundant primers of Xci-specific detection.

Figure 3. Specific detection of Xanthomonas campestris pv. incanae strain from other closely related Xanthomonas campestris pathovars such as pv. campestris, pv. raphani and other bacterial strains using the Xci specific developed primers (XCI_1F/R, XCI_2F/R, XCI_3F/R, XCI_5F/R and XCI_6F/R) in PCR based assay. 100 bp DNA ladder; Lane 1: Xci (HRI-W-6377); Lane 2-9: Xcc (HRI-W-3811, HRI-W-3849A, HRI-W-5212, HRI-W-1279A, HRI-W-3880, HRI-W-6181, HRI-W-8450A and MBG145.3); Lane 10: Xcr (HRI-W-8305) Lane 11: Xev (KACC11153); Lane 12: Xag (KACC10491); Lane 13: Xcz (KACC17126); Lane 14: Xc (KACC10490); Lane15: Pseudomonas syringae pv. maculicola (ICMP13051); Lane16: Erwinia carotovora subsp. carotovora (ICMP12464); Lane 17: Plasmodiophora brassicae; Lane 18: Didymella bryoniae (NIHHS1326) and Lane 19: Negative control (distilled water).

We identified the orthologous genes and their genomic neighborhood by EDGAR tool. In addition, we identified only the 5 orthologs genes of Xci rather than any other pathovars of Xanthomonas genomes. Aparant locations and orientations of the 5 selected genes on Xanthomonas genome are shown in the Table 3. Among those genes, two genes are located on positive strand (+) (XCICFBP2527_ RS13555 and XCICFBP2527_RS21720) and other three on negative strand (-) (XCICFBP2527_RS16430, XCICFB P2527_RS03935 and XCICFBP2527_RS21980) (Table 3).

Detection limits of primers in Xci genomic DNA concentration

A PCR sensitivity test for Xci specific primers was performed using genomic DNA of Xci strain (HRI-W-6377). The sensitivity of the developed primers was checked using a ten-fold dilution series of genomic DNA. Five primer sets; XCI_1F/R, XCI_2F/R, XCI_3F/R, XCI_5F/R and XCI_6F/R gave effective sensitivity in PCR assay. The expected PCR product was amplified with each primer up to sixth dilution (0.00007 ng/µL) with the exception of one primer pair XCI_2F/R, which resulted amplification up to fifth dilution (0.0007 ng/µL)

(Fig. 4) and indicated detection limit by PCR.

Figure 4. PCR detection limits regards to the genomic DNA concentration of Xci strain (HRI-W-6377) using developed markers (XCI_1F/R, XCI_2F/R, XCI_3F/R, XCI_5F/R and XCI_6F/R) in PCR based assay.

Comparative genomics is an easy and accessible approach, which provides an opportunity for in-depth analysis of intr-aspecies diversity by using the available genome sequences of different strains in the public data bases. Pangenome of bacterial populations has been used since 10 years ago for comparative genomics; currently, several bioinformatics tools have been developed for comparative genome analysis. However, a fewer online tools is available for comparative genome analysis, especially for bacterial genomes. Among them, EDGAR 3.0 is one of the most popular webserver based tool for the comparative bacterial genome study (Blom et al. 2009; Blom et al. 2016; Yu et al. 2017; Dieckmann et al. 2021). This tool could be used successfully in genomic subset calculation, construction of Venn diagram or synteny plots (Supplementary Fig. S1) in terms of robustness, user friendliness and graphical interpretation of the obtained results. The evolutionary relationships among the closely related bacterial genera Erwinia, Tatumella and Pantoea as well as the species within Pantoea have been recently done by using EDGAR 2.0 software (Palmer et al. 2017). The genetic basis and phylogenetic distribution of the insecticidal traits of P. fluorescens against herbivorous insects has also been reported by using the same software (Flury et al. 2016). EDGAR has resolved the structure of the heterogenous actinobacterial taxon of Rhodococcus to identify the boundaries for making intrageneric and supra-specific classification (Sangal et al. 2016). The similar approach was found effective to assess the structure of the genus Geobacillus (Aliyu et al. 2016). This tool provides a deep understanding on the whole genome information, especially in the aspects of core genome, dispensable genome and singletons for better interpretation of phylogenetic relationships and species diversity. This approach offers opportunity to develop simple, quick, sensitive and accurate PCR based detection technique compared to other techniques. Herein, we aimed in differentiating the genome content of Xci from a pool of bacterial population for developing Xci specific molecular markers in PCR detection (Fig. 2).

Several molecular methods are used in detection of closely related pathogenic bacterial strains that affecting various crops (Berg et al. 2005). Xci specific markers have been yet to develop for diagnosing its’ infection in field grown crucifers. In this regards, 10 singleton genes were randomly selected from 82 retrieved genes for developing Xci specific markers including one T3E gene for designing 11 PCR based primers. Five primer pairs were produced amplicons specific to Xci strain (Fig. 3, Supplementary Table S1).

T3E protein, a crucial virulence factors in Xci bacteria, which can be disrupted host cell signaling, cytoskeletal rearrangement, vesicular traffic as well as autophagy in the process of disease development. Bacteria are penetrated into the host cell via a specialized injection apparatus called T3E secretion system and mimic host defense molecules (Abe et al. 2005). Hoobs et al. (2016) have been used Genome Search Effector Tool (GenSET) for mining species specific effectors to study the different translocation signals across different T3E gene families. Currently, researches are concentrated on the prediction and analysis of T3E gene to understand the bacterial pathogenicity (Goldberg et al. 2016). T3E genes are believed to confer specificity in Xci for pathogenicity in ornamental crucifers. Except one T3E gene, the remaining 81 genes are encoded hypothetical proteins might also play role in species specificity. Thus, out of five well amplified Xci specific primers, four are situated on genes of hypothetical proteins and could also be used effectively in discriminating Xci from a mixture of bacterial populations.

Four primer pairs (XCI_1F/R, XCI_3F/R, XCI_5F/R and XCI_6F/R) were able to detect genomic DNA at a low limit of 0.00007 ng/µL (Fig. 4) which an agreement with the dilution test in detecting Xcc (Singh et al. 2014). For understanding the effective detection limit, it is important to diagnose the presence of bacteria in the fields at asymptomic stage. We claim that our primers are very effective and sensitive to detect the Xci strain even at very low DNA concentration. This is the first report to develop Xci-specific markers through comparative genomics approach using EDGAR.

Bacterial blight disease is an important disease of many crops or plants which is caused by Xci pathogen but there is no marker is available for detection. Therefore, PCR based markers could be used to detect Xci in plant quarantine. However, the efficiency of Xci specific PCR based marker was tested in eighteen bacerial species, but test experiment could be broaden by including more Xanthomonas species upon availability.

The authors have all ethical responsibilities of this manuscript. This research has conducted without any commercial or financial benefits, the authors also declared no conflict of interest and none of the animal and/or human trials were included.

We thank Joana G. Vicente, University of Warwick, UK for providing different strains with known races of Xcc, Xci and Xcr isolates; also thank to Dr. Pilar Soengas for giving Xcc race 8 strain and the Korean Agriculture Culture Collection (KACC), Korea and the International Collection of Microorganisms from Plants (ICMP), New Zealand for providing respective bacterial strains.

I-SN, J-IP and H-TK conceived and designed the study. DMIJ conducted the in silico analysis. MHR prepared and cultured the samples, isolated DNA, performed wet lab validation and wrote the manuscript. UKN extensively revised the manuscript and edited it. HJJ prepared the bacterial inocula and managed the field infected samples. All authors read and approved the final draft of the manuscript.

This work was supported by the Center for Horticultural Seed Development (Golden Seed Project No. 213007-05-5-SB510) of the Ministry of Agriculture, Food and Rural Affairs (MAFRA), Republic of Korea.