Genetic diversity of Pantoea stewartii subspecies stewartii causing jackfruit-bronzing disease in Malaysia

Jackfruit-bronzing is caused by bacteria Pantoea stewartii subspecies stewartii (P. stewartii subsp. stewartii), showing symptoms of yellowish-orange to reddish discolouration and rusty specks on pulps and rags of jackfruit. Twenty-eight pure bacterial strains were collected from four different jackfruit outbreak collection areas in Peninsular Malaysia (Jenderam, Maran, Muadzam Shah and Ipoh). Positive P. stewartii subsp. stewartii verification obtained in the study was based on the phenotypic, hypersensitivity, pathogenicity and molecular tests. Multilocus sequence analysis (MLSA) was performed using four housekeeping genes (gyrB, rpoB, atpD and infB) on all 28 bacterial strains. Single gyrB, rpoB, atpD and infB phylogenetic trees analyses revealed the bootstrap value of 99–100% between our bacterial strains with P. stewartii subsp. stewartii reference strains and P. stewartii subsp. indologenes reference strains. On the other hand, phylogenetic tree of the concatenated sequences of the four housekeeping genes revealed that our 28 bacterial strains were more closely related to P. stewartii subsp. stewartii (99% similarities) compared to its close relative P. stewartii subsp. indologenes, although sequence similarity between these two subspecies were up to 100%. All the strains collected from the four collection areas clustered together, pointing to no variation among the bacterial strains. This study improves our understanding and provided new insight on the genetic diversity of P. stewartii subsp. stewartii associated with jackfruit-bronzing in Malaysia.


Introduction
Jackfruit (Artocarpus heterophyllus) comes from the mulberry family (Moraceae) and is believed to originate from the evergreen forest in Western Ghast, India [1]. Jackfruit is an

Sample collection and isolation of bacterial pathogen
A total of 28 diseased jackfruits (Tekam Yellow J33 variety) were collected from outbreak plantation areas in Peninsular Malaysia, comprising of 7 fruits from Jenderam (Selangor), 7 fruits from Maran (Pahang), 7 fruits from Muadzam Shah (Pahang) and 7 fruits from Ipoh (Perak) from September-November 2017 (Table 1). Fruits were collected randomly, with three replicates for each sample per location (total 21 fruits per collection area). The infected fruits were wrapped, put inside sterile plastic bags, labelled and brought to the laboratory for initial diagnosis and isolation of the pathogen. The isolation procedure was conducted not later than 24 h after the collected samples were brought to the laboratory. Infected samples were excised, cut and rinsed with sterilized distilled water (containing 1% of sodium hypochlorite, Chlorox™) for 3 minutes. The jackfruit pulps were put on King's B agar medium, sealed properly and incubated in inverted position for 24 to 48 h at 28º C. After incubation, isolated colonies were subcultured until pure colonies of suspected P. stewartii subsp. stewartii strains were obtained. As mentioned by EPPO, the colonies were expected to be lemon to pale-yellow or orange-yellow in colour, flat to convex, transparent with entire edges, and slow to medium growing [9]. For later use, the pure bacterial cultures were grown in nutrient broth medium with 20% (v/v) glycerol and stored at -80˚C.

Ethics statement
For collection of fruits, no specific permits or permission were required for these locations/ activities. All field research was conducted on privately-owned farms and with owners' permission. All locations where the fruits were collected did not involve endangered or protected species.

Verification via phenotypic characterization
The positive phenotypic characterization for the verification of P. stewartii subsp. stewartii were performed based on previous studies verification [5,9,[17][18][19] as mentioned on S1 Table. All tests were performed using the freshly grown colonies of P. stewartii subsp. stewartii (24 to 48 h).

Hypersensitive assay
Four-week-old Nicotiana benthamiana transplants were used for hypersensitivity assays. Using syringe, 1 ml of 10 8 CFU/ml of bacterial suspension (as determined by McFarland Standard, BioMérieux, Marcyl'Etoile, France) was inoculated into a fully expanded tobacco leaf (3 leaves per bacterial culture). Leaves infiltrated with sterile distilled water served as a negative control. The greenhouse condition maintained at temperature range from 26˚C to 35˚C. The plant response for Hypersensitivity reaction (HR) was recorded after 10 to 36 h post-infiltration. Necrosis of the infiltrated tissues was considered as positive HR, and the absence was considered as a negative test. Four replicates were performed for each strain due to any imprecision concern. Two results were expected for this test, as past studies showed P. stewartii subsp. stewartii strains resulted in positive HR and pathogenic [18] or otherwise resulting negative HR, suggesting that the strains are non-pathogenic due to mutation in the hrp gene [5,11,18].

Pathogenicity test
Detached healthy jackfruit pulps (Tekam Yellow J33 variety) were inoculated with 10 mL of 10 8 CFU/mL (as determined by McFarland Standard, BioMérieux, Marcyl'Etoile, France) of 24 to 48 h pure P. stewartii subsp. stewartii suspensions. Sterile distilled water was inoculated to control healthy jackfruit pulps. 4 replicates were performed per strain. All 28 bacterial strains treatments were put inside sterile petri dish (properly sealed) and incubated at 28 ºC in a controlled chamber. The evaluation of jackfruit-bronzing symptoms was recorded daily up to two weeks of inoculation.

MLSA
All extracted total genomic DNA were used as a template in the PCR amplification using primers as listed on S2 and S3 Tables. The PCR reactions were performed as described under the previous section.

Electrophoresis of PCR products, DNA purification, sequencing, sequence alignment, phylogenetic analyses and dendrogram construction
Separation of PCR products was performed on a prepared 1% agarose gel with FloroSafe DNA stain. Using Mini Sub™ DNA Cell, (Bio-Rad Laboratories Inc., USA), the PCR amplicons was run at 75 V (Power supply model 1000/500, Bio-Rad Laboratories Inc., USA) for 45 min and later visualized and photographed by Alpha Imager System (Alpha-Innotech, Siber-Hegner, UK). All amplified PCR products were then sent for purification and sequencing (MyTACG Sequencing, Malaysia). All forward and reverse sequences were assembled and analyzed using Bioedit version 7.2.5 [21]. The FASTA format of the gene sequences were subjected to BLASTn (http://blast.st-va.ncbi.nlm.nih.gov/Blast.cgi) search to determine their similarity to nucleotide sequences available in the GenBank database. This was followed by using ClustalW multiple alignment program (http://www.ebi.ac.uk/Tools/msa/clustalw2/), with manual adjustments. MEGA7 software [22] was used for the construction of phylogenetic trees using the best fit substitution model identified and calculated by the "Find Best Models" tool. Maximum likelihood tree were then generated with bootstraps values of 1000 for each analysis. The available sequences for these genes in genomes of P. stewartii subsp. stewartii were then compared.
All reference strains were obtained from NCBI online database (S4 and S5 Tables). The sequenced genes from the study have been deposited into the NCBI database as mention on S4 and S5 Tables.

Collection of the sample and symptoms of jackfruit-bronzing
From the total of 28 collected jackfruits, all fruits (7 fruits from Jenderam, 7 fruits from Maran, 7 fruits from Muadzam Shah and 7 fruits from Ipoh) showed the symptoms of bronzing disease ( Table 2). The outer appearance of the collected fruits were symptomless. Internally, they showed visible jackfruit-bronzing symptoms either yellowish discolouration with bronzing specks or reddish discolouration on the pulps and rags (Fig 1). The symptoms were inconsistent with 17 of the fruits showed yellowish discolouration with bronzning specks and 11 had reddish discolouration appearance ( Table 2).

Phenotypic characterization and molecular characterization
All the colonies were identified accordingly to the morphology characteristics as mentioned by EPPO [9]. Colonies were lemon-yellow in colour, circular, raised and convex on King's B agar medium (Fig 2). From the 28 strains, all of the strains were positive and verified for P. stewartii subsp. stewartii identification (Table 3). PCR amplification of all 28 isolated strains using 16S-23S ITS primer pair, each produced a~900 bp amplicon (S1 Fig). Phylogenetic analysis based on partial 16S-23S ITS gene (Fig 3) revealed all strains isolated in this study were 100% clustered together with the P. stewartii subsp. stewartii reference strains DC283, S3 and W1. None of the strains clustered with P. agglomerans (Pantoea agglomerans) and P. ananatis (Pantoea anantis) strains.

Hypersensitive assay
Hypersensitivity response (HR) induced by all 28 strains at 24 h after inoculation is shown in

Pathogenicity test
All 28 P. stewartii subsp. stewartii strains produced jackfruit-bronzing disease symptoms within 14 days post-inoculation (Fig 5). No disease symptom appeared on the control. The reisolation of the inoculated strains were tested further via phenotypic and molecular characterization and showed similar characteristics to positive P. stewartii subsp. stewartii test.

Discussion
Systematic identification and characterization were performed for the verification of P. stewartii subsp. stewartii strains from the 4 different jackfruit-bronzing disease outbreak collection area in Peninsular Malaysia. The strains were verified by morphological, phenotypic and molecular characterization test listed by (9). Starting with the isolation and phenotypic characterization, phenotypic tests were performed. Our study showed all 28 strains demonstrated a similar and parallel response to all phenotypic tests and resembled P. stewartii subsp. stewartii characteristics [5,9,[17][18][19]. Motility test is crucial because stated that most bacteria under Pantoea genus are motile due to the presence of peritrichous flagella except for P. stewartii subsp. stewartii [11,23,24]. Indole test is also important in subspecies identification [18,25], as P. stewartii subsp. indolegenes distinguished and identified from P. stewartii subsp. stewartii through the positive result of indole test with Kovac's reagent. So, the outcome of our indole and motility test results from the 28 strains showed strong positivity towards the identification of P. stewartii subsp. stewartii. HR on tobacco (Nicotiana tabacum) was performed on P. stewartii subsp. stewartii and shown to be necrotic [14,18,26]. P. stewartii subsp. stewartii contains and carries a specialized protein secretion complex known as type III secretion systems (T3SSs) producing needle-like injectisomes for the delivery of effector proteins into host cells [11,27]. A previous study by Correa et al. (2012) confirmed that P. stewartii subsp. stewartii obtained from maize contains Hrc-Hrp1 as well as Inv-Mxi-Spa type T3SS, which were responsible for the pathogen colonization of susceptible hosts, and transmission from the vector to host plants by the transversal of the thick plant cell wall through the generation of thin pili. This hrp gene (resistant genes) secretion of P. stewartii subsp. stewartii is required to elicit a HR in tobacco leaves [14,28,29]. HR involves rapid reaction, and sometimes it only takes 8-12 h after being inoculated with the bacterial pathogen to become necrotic and eventually die [29]. Our hypersensitivity assay result was similar with previous reports [14,18], T3SSs and hrp genes were believed to promote parasitism and leads to pathogenesis, so other than virulence, tobacco hypersensitivity is also a very convenient way to screen bacterial cultures for pathogenicity [11,30].
Further confirmation was obtained from molecular characterization analysis. Standard differentiation of bacterial species using 16S-23S ITS region was believed to produce better sequence variability than the 16S rDNA gene as 16S rDNA gene cannot be separated easily by using electrophoresis [20]. Primer pair ES16/ESIG2c was tested on P. stewartii (33 strains), other Pantoea spp. (12 strains), related Erwinia spp. (14 strains) and Xanthomonas campestris pv. campestris (1 strain) and showed that only P. stewartii strains and one P. ananas (P. ananatis) strain yielded at the expected amplicons (0.92-kb fragment) [14]. This primer pair also confirmed the jackfruit-bronzing disease from Muadzam Shah plantation of Pahang state was caused by P. stewartii subsp. stewartii [6]. Our study has proven that all our 28 strains yielded at the expected amplicon's size and sequence similarity of 100% to other P. stewartii subsp. stewartii strains obtained from GenBank database. Phylogenetic analysis constructed using  MEGA7 on the ES16/ES1G2c gene sequences showed our strains clustered with the P. stewartii subsp. stewartii reference strains and none of the strains clustered to P. agglomerans and P. ananatis strains. For the first time, genetic diversity of P. stewartii subsp. stewartii strains causing jackfruitbronzing disease in Malaysia has been studied via MLSA. MLSA is a genotypic characterization method using the sequences of multiple protein-coding genes of prokaryotes for taxonomic purposes [31]. The stable housekeeping genes that codes for functional proteins make MLSA as a powerful method with a rapid genetic modification in comparison to 16S rDNA/ rRNA [31]. Previous genetic diversity studies using 16S rDNA/rRNA reported the clustering and close relationship of P. stewartii subsp. stewartii with P. ananatis and P. agglomerans [32,33], P. stewartii with P. ananatis [34], P. stewartii subsp. stewartii with Erwinia stewartii [35] and P. stewartii subsp. stewartii with P. ananatis, Pantoea deleyi, P. agglomerans, Pantoea vagans, Pantoea eucalypti [36]. Undoubtedly, MLSA overpowers 16S rDNA/rRNA in taxanomic discrepancies in the identification purposes of Pantoea genus [37][38][39]. Four housekeeping genes were used in this study, as recommended by previous studies on MLSA [40][41][42][43][44]. All phylogenetic analyses in this study were performed with bootstrap replication of 1000, as it supports higher phylogenetic differentiation of Pantoea species [40,45]. The three outgroups were previously known as 'Japanese species' of Pantoea, namely Pantoea citrea, Pantoea punctata and Pantoea terrea [40]. However, MLSA indicated that the taxonomic position of these  "Japanese species" were clustered together with Tatumella ptyseos, so they were then transferred to the genus Tatumella. Phylogenetic trees obtained from MLSA based on sequences of gyrB, rpoB, atpD and infB confirmed the allocation of all our strains to the genus Pantoea. Phylogenetic analysis showed that all of the 28 bacterial strains collected from the four collection areas were clustered together and phylogenetically most closely related (bootstrap value of 99-100%) to P. stewartii subsp. stewartii and P. stewartii subsp. indolegenes reference strains. Our result is congruent with the previous studies of MLSA performed on Pantoea species where phylogenetic trees based on housekeeping genes (gyrB, rpoB, atpD and infB) revealed 99-100% similarities between P. stewartii subsp. stewartii and P. stewartii subsp. indolegenes [35,40,[45][46][47][48][49][50][51].
However, MLSA based on the concatenated genes were able to distinguish that our strains were more closely related to P. stewartii subsp. stewartii (99% similarities) from its close relative P. stewartii subsp. indologenes, although sequence similarity between the subspecies was high (up to 100%). Study of MLSA that performed less than four housekeeping genes were insufficient and more bias, due to low-resolution power or genes impaired [44]. High discriminatory power is gained from MLSA from the concatenated genes than single gene [52]. In other words, the number of genes selection must be considered to achieve more discriminatory power [52].
All bacterial strains collected from the four areas of Malaysia were clustered together with the reference strains, indicating that the strains were genetically similar and share close phylogenetic relatedness with each other. This emphasizes that all our strains were likely originated and derived from a single emergence a long time ago [53]. Based on phylogenetic tree of the concatenated sequences of the four housekeeping genes (Fig 7), no major changes in the clustering pattern appeared since the position of the P. stewartii subsp. indologenes reference strains were still in one cluster (100% similarity) with our bacterial strains and P. stewartii subsp. stewartii reference strains. Hence, in our case, all the 28 P. stewartii subsp. stewartii bacterial strains (single and concatenated) might have a common origin with P. stewartii subsp. stewartii collected from USA (LMG 2713, LMG 2715 and LMG 2718) and P. stewartii subsp. indologenes bacterial strains collected from USA, Hawaii and India (PNA 14-12, LMG 2673 and LMG 2631).
In conclusion, our study verified and confirmed that bacteria causing jackfruit-bronzing in four areas of Malaysia is P. stewartii subsp. stewartii. We highlighted the genetic diversity of P. stewartii subsp. stewartii associated with jackfruit-bronzing in Malaysia, providing new insight on no genetic variation among the collected strains. Phylogenetic analyses indicated that our strains were closely related to P. stewartii subsp. stewartii strains (concatenated genes) and both P. stewartii subsp. stewartii and P. stewartii subsp. indologenes reference strains (single genes). As the improvement in jackfruit industries in Malaysia is vital, once jackfruit-bronzing disease is properly investigated, management options can be deployed to mitigate the disease impact. Controlling plant disease involves proper verification of the causal agent, where in this study, we provide information on the phenotypic tests, hypersensitivity test, pathogenicity test, molecular test, as well as genetic diversity evaluation through MLSA of P. stewartii subsp. stewartii. The output and data of this study will be useful as a proper reference and documentation in quarantine purposes and prevention of the spread of jackfruit-bronzing disease and its causal pathogen, P. stewartii subsp. stewartii. Supporting information S1