MATERIALS AND METHODS

The investigation was carried out in three phases. The first consisted in the isolation, morphological characterization, physiological and molecular tests to identify the bacterial isolation. The second and third consisted in the inoculation of different plant structures and bean plants (Phaseolus vulgaris) under controlled conditions to determine its infective potential.

Biological material. We analyzed bulbs from white onions (Allium cepa) of the Carta blanca variety with soft rot symptoms from the town of San Vicente Chicoloapan, State of Mexico, located on latitude N 19° 24’ 49” and longitude W 98° 53’ 8.29”.

Bulbs from red onion (A. cepa), leek (A. ampeloprasum) stalks, cactus (Opuntia indica) cladodes, carrot (Daucus carota) roots, ginger (Zingiber officinale) rhizomes, broccoli (Brassica oleracea) inflorescences, mushroom (Agaricus bisporus) sporocarps and bean (Phaseolus vulgaris) seeds were also used to evaluate the ability of colonization and the pathogenicity of the bacterial isolation.

Isolation and morphological characterization of the phytopathogen. The external cataphylls of onion bulbs with signs of soft rotting were cut longitudinally. The bacterial flow was observed using distilled water under a light microscope at 10X. A bacterial sample was taken, then planted in King’s B medium and incubated at 28 °C for 48 h. The bacterial culture was morphologically characterized under the stereoscopic microscope. Afterwards, hypersensitivity tests were performed on tobacco plants (Nicotiana tabacum “xanthi”) aged four months, as well as on the rotting of potato tubers (Solanum tuberosum) (^{Goszczynska et al., 2000}).

Koch Postulates. The isolated strain was inoculated on healthy tissue from white onion, variety Carta blanca, with the application of 0.5 mL of an aqueous suspension with 3X10⁸ UFC mL^-1, adjusting it using ^{McFarland’s scale (McFarland, 1907}). As controls, four slices of healthy white onion were inoculated with 0.5 mL of sterilized distilled water and placed in a wet chamber at 28 °C. The rotting of the cataphylls was evaluated 48 h after inoculation. A sample of the exudate of the tissue with rotting was planted in King’s B medium for 48 h. Later, the bacterial growth was observed under the stereoscopic microscope, and a bacterial colony grown in isolation was selected and transferred to culture media King’s B, MacConkey, and Congo Red (^{Rodríguez,1982}).

Physiological and biochemical characterization. The following physiological and biochemical tests were carried out: Gram stain, oxidase, catalase, KOH reaction, nitrate reduction, oxidative-fermentative metabolism, gel liquefaction, arginine hydrolysis, levana production, starch hydrolysis and use of carbon sources following the methodology described by ^{Shaad (2001}). The bacterial growth was registered in a pH of 4 to 10 (^{Estrada de los Santos et al., 2013}) and temperatures of 30 to 43 °C (^{Kowalska et al., 2015}).

DNA extraction and polymerase chain reaction (PCR). Bacterial DNA extraction was carried out using the CTAB method quoted by ^{Doyle and Doyle (1990}), modified by washing with con sodium acetate 3 M. The quantity and quality of the DNA was evaluated by spectrophotometry in Nanodrop 126 (ND-1000, Nanodrop Technologies).

For the amplification of the gene 16S rRNA, the universal primers 8 F (5´AGAGTTTGATCCTGGCTCAG-3´) and 1492 R (5´GGTTACCTTGTTACGACTT-3´) (^{Galkiewicz and Kellogg, 2008}) were used. The amplification of gene 16S rRNA was carried out under the following conditions, in a final reaction volume of 25 µL: an initial DNA denaturalization was carried out at 95 °C for 5 min, followed by 30 cycles at 94 °C for 1 min, 54 °C for 45 s, 72 °C for 1 min and a final extension at 72 °C for 8 min. The amplified fragment was separated by electrophoresis in 2% agarose gel, with ethidium bromide at 85 volts for 60 min. The resulting fragment was observed under ultraviolet light in a Biologing Systems photodocumenter, model Epi Chemi II Darkroom. Later, the product was sequenced by the Macrogen company in Korea.

Phylogenetic tree. To confirm the identity of the bacteria, the assembly of the forward and reverse sequences were assembled using the software MEGA X (^{Kumar et al., 2018}). The consensus sequence was compared with the sequences registered in the Basic Local Alignment Search Tools (BLAST) of the National Center for Biotechnology Information (NCBI). To create the phylogenetic tree, the partial sequences of gene 16S rARN were discharged from the GenBank from B. gladioli (MN559413), B. cepacia (AF097530), B. ambifaria (AF043302), B. plantarii (AB183679) and B. glumae (NR_029211), Paraburkholderia graminis (NR_029213), Pseudomonas viridiflava (NR_117825), P. fluorescens (NR_113647) and P. aeruginosa (NR_113599), using Agrobacterium tumefaciens (NR_115516) as an out-of-group species. The multiple alignment of sequences was carried out with the MUSCLE option. The phylogenetic tree was obtained using the method of Maximum Likelihood (ML) with 500 bootstrap repetitions, based on the Tamura-Nei model (TN93+G). The model used was obtained using a model test, and according to the Bayesian Information Criterion (BIC), the one that best describes the base substitution pattern was chosen.

Scanning electron microscopy. A colony from the bacterial isolation was used, with 72 h growth in King’s B medium at 28 °C for the observation of the culture’s morphology under a scanning electron microscope. A small bacterial sample was taken and fixed in 3% glutaraldehyde. It was then rinsed with Sorensen’s phosphate buffer (0.1 M), at a pH of 7.2 for 10 min. After 24 h, this procedure was carried out twice. It was then dehydrated with alcohol at concentrations of 30, 40, 50, 60, 70, 80, 90 and 100%. It was left to rest in each concentration for 30 min. The process was only repeated with a concentration of 100%. Later, the sample was dried until it reached the critical point using the Samdri-780A^® drier (U.S.A., 2007). Next, the specimen was placed on a slide using a brush under a stereoscopic microscope. The sample was then covered in gold using the FINE COAT (Ion sputter JFC-1100) equipment. Finally, the bacterial morphology was observed with the AXS Microanalysis software for the scanning electron microscope (Oxford instruments INCA x-act 2009).

Inoculation of the bacterial isolation in plant structures under laboratory conditions. Red onion (A. cepa) bulbs, leek (A. ampeloprasum) stalks, cactus (O. indica) cladodes, carrot (D. carota) roots, ginger (Z. officinale) rhizomes, broccoli (B. oleracea var. italica) inflorescences and mushroom (A. bisporus) sporocarps were disinfested using sodium hypochlorite (NaClO) at 2% for 2 min, and then with 70% ethanol for 1 min, followed by three washings with sterile distilled water. Finally, they were dried on sterilized paper towels for 10 minutes. Then, they were injected with 0.5 mL of bacterial inoculant at a density of 3.0 X10⁸ UFC, adjusted using McFarland’s scale. In the controls, the bacterial inoculant was substituted with sterile distilled water. Afterwards, the plant structures were placed in humid chambers and the damage was evaluated after 48 h. To determine the severity, a scale of 1 to 5 was created for the degree of maceration: 0, 20, 40, 60, 80 and 100%, respectively (^{Liao et al., 1986}). A completely random design was used, with four repetitions per treatment. With the data obtained, an ANOVA was carried out, and the means of the treatments were then compared using Tukey’s test (p≤0.05). The program used was SAS, Version 9.0 (^{SAS Institute, 2002}).

Inoculation of bean with the bacterial isolation under greenhouse conditions. Each bean seed was germinated in a 1 L container with Peat Moss and Agrolite (75:25). A bacterial suspension of the pathogen was inoculated at a concentration of 3.0 X10⁹ UFC in four plants, 10 days after germination; likewise, four plants were left as negative controls (^{Falcao et al., 2004}). A completely random design was used, with four repetitions per treatment. With the data obtained, an ANOVA was carried out, and the means of the treatments were then compared using Tukey’s test (p≤0.05). The SAS program, version 9.0 (^{SAS Institute, 2002}) was used.

In vitro sensitivity to chemical and biological products. A bacterial isolation sensitivity test was carried out using the method of diffusion in agar (^{Matuschek et al., 2014}), on the following chemical products: agricultural cuprimicin (200 g/ 100 L), serenade powder (1 Kg/200 L), bactrol 2X (60 g/100 L), agricultural bactriomicin (400 g/100 L), biological fungifree (2.5 Kg/200 L), cuprimicin 500 (625 g/100 L), phyton (1.5 mL/1 L), copper oxychloride (400 g/100 L), kasumin (2 L/200 L), final bacter (1.6 Kg/200 L), quatz (400 g/100 L), biological bacterbest (1 L/200 L) and biological biotermin (3 L/200 L). As a control, we used sterile distilled water. We placed 200 µL of an aqueous solution of the bacterial isolation with 3.0 X10⁸ UFC in squared Petri dishes (120 X 120 mm) with a nutritious agar medium. The inoculant was distributed evenly with a Driglasky spatula on the surface of the medium. Sterile filter paper discs (13 mm in diameter) were used, which were submerged for three seconds in each of the products and placed in the medium dishes. Finally, the dishes were incubated at 28 °C for 48 h to look for the presence or absence of inhibition areas (^{Bauer et al., 1966}).

RESULTS AND DISCUSSION

Morphological characterization. The morphology of the cultures observed in King’s B medium, after 48 h of incubation at 28 °C, was smooth, yellow, convex, mucoid cultures, with a production of diffusible pigment in the center (Figure 1A). These characteristics coincided with those reported by ^{Lamovsek et al. (2016}) for B. gladioli pv. allicola. However, the cultures presented no fluorescence in King’s B medium, as reported by ^{Kowalska et al. (2015}).

The cultures were purple in color in MacConkey’s agar medium (Figura 1B), which confirmed that this bacterium is able to ferment lactose, unlike Pseudomona aeruginosa (^{Callicó et al., 2004}), which lacks lactase (^{Ranjan et al., 2017}). On the other hand, the bacterial isolation displayed a growth of red cultures in a Congo Red medium (Figure 1C), which indicated that this bacterium can assimilate the iron found in the culture medium. This characteristic is crucial to the colonization process of the phytopathogenic microorganisms (^{Aguado et al., 2012}).

Molecular identification. The amplification of gene rRNA of the bacterial isolation displayed fragments of 1500 pairs of bases (pb). When comparing the consensus sequence (Accession Number MT672591) on the NCBI data base, there was a similarity of 100% with B. gladioli.

Phylogenetic tree. The consensus sequence MT672591 was grouped with B. gladioli (MN559413) with a similarity of 100 % (Figure 2).

Pathogenicity. The onion slices displayed soft rotting 24 h after inoculation. The stereoscopic microscope showed abundant bacterial flow 48 h after inoculation at 28 °C. B. gladioli (MT672591) was positive to the hypersensitivity reaction in tobacco and rotting in potato. The control slices presented no symptoms of soft rotting.

Scanning electron microscopy. The results of the scanning electron microscope coincide with reports by the ^{Boston Research Occupational Health Program (2019)}, which indicate that members of the Burkholderia genus present a bacillary shape, between 1.6 and 3.2 µm, with one or several flagellar poles, important to the pathogenicity (^{Tomich et al., 2002}). In the present study, it was not possible to visualize the flagella, since the observation of these structures is carried out with negative staining with a transmission electron microscope (^{Jurado and Petruccelli, 2005}).

Figure 1 Growth of the B. gladioli bacterial isolation in King’s B culture medium (A), MacConkey’s Agar (B) and Congo Red (C) after 48 h of incubation. 

The B. gladioli bacilli (MT672591) formed biofilms and were also observed in a planktonic state (Figure 3). The formation of biofilms is another important characteristic in this genus, since it is through them that they detect signaling molecules, helping their synchronization in the gene expression (^{Federle et al., 2009}).

Physiological and biochemical characterization. B. gladioli (MT672591) was positive to the hypersensitivity reaction test in tobacco and potato rotting, which confirms the presence of secretion systems I, II and III (^{Seo et al., 2015}).

Figure 2 Phylogenetic tree based on the comparison of the gene sequences of subunit 16S of the rRNA of species in the Burkholderia and Pseudomonas genera. The parentheses indicate the accession numbers of the sequences in the NCBI data base. 

In the phytopathogenic bacteria, Secretion System III is codified by the HRP genes (hypersensitive response and pathogenicity), which are necessary for the bacteria to cause diseases in susceptible plants and cause a hypersensitivity response in resistant plants (^{Lindgren et al., 1986}; ^{Büttner and He, 2009}).

These secretion systems help the bacteria to colonize different ecological niches, such as postharvest fruits and in vivo plant tissues with the hosts, with which a compatible relationship is established. B. gladioli (MT672591) was strictly aerobic and used glucose, sucrose, arginine, and lactose as a source of carbon, coinciding with reports by ^{Kowalska et al. (2015}). The bacterium (MT672591) grew to a pH of 4, which gives it a competitive advantage in acid soils. This contrasts with reports by ^{Estrada de los Santos et al. (2013}), who mention that none of the 59 strains evaluated, belonging to 43 species of the Burkholderia genus, grew to a pH of 4. Growth of B. gladioli (MT672591) was observed at 42 °C, agreeing with reports by Kowalska et al. (2015). Table 1 shows the similarities and differences between the isolated strain in this study and those reported in Sinaloa, Mexico, and

Figure 3 Micrography of B. gladioli (MT672591) observed with a scanning electron microscope. A and B correspond to biofilms (5,000 X and 10,000 X, respectively); C and D correspond to the bacteria in a planktonic state (5,000 and 20,000 X, respectively). 

Inoculation of B. gladioli in plant structures under controlled conditions. Significant differences were registered in the degree of maceration (Table 2). The highest degree of maceration was found in the red onion bulb and the leek stalk. The tissues of both plant structures were completely macerated 48 h after inoculation and it displayed the typical symptom of soft rot, caused by B. gladioli (MT672591) (Figures 4 A and B). The red onion and the leek presented similar interactions with the bacterium, probably due to them both belonging to the Alliaceae family.

In the cactus (Opuntia spp.) cladodes, the first symptoms of soft rot were observed 24 h after inoculation. First, dark brown dots appeared, which grew with a moist aspect and a halo around the spots 48 h after inoculation (Figure 4C). A complete rotting of the cladode was observed six days after inoculation, coinciding with the symptoms caused by Pectobacterium spp. (^{Torres et al., 2016}). Both phytopathogens are characterized for having high enzymatic activity. To this day, there are no reports on the Burkholderia genus as a pathogen of cactus in a natural state; however, due to the aggressiveness in the colonization observed under controlled conditions, it deserves further studies.

Likewise, soft rot and degradation of the sporocarp were observed in mushrooms (Figure 4D). ^{Chowdhury and Heinemann (2006}) reported that Secretion System type II is involved in the disease caused by B. gladioli pv. agaricola in mushrooms. These researchers found that avirulent mutants were unable to degrade the tissue of mushrooms, due to the reduction in the ability to secrete chitinases and proteases, as well as the reduction in the number of flagella. B. gladioli pv. agaricola is considered a potential mushroom pathogen, which could cause important losses in the industry of this fungus (^{Gill and Tsuneda, 1997}).

Figure 4 Symptoms of rotting caused by B. gladioli (MT672591) in different plant structures under controlled conditions: A) red onion bulbs, B) leek stalks, C) cactus cladodes, D) mushroom sporocarps, E) carrot roots and F) broccoli inflorescences. 

In carrots, dark brown patches were produced near the points in which the bacterium was inoculated, three days after inoculation (Figure 4E). So far, there are no reports on damages to carrots by B. gladioli, although Pseudomonas viridiflava has been reported to promote similar symptoms in carrot roots (^{Almeida et al., 2013}). This similarity may be due to the proximity of the Burkholderia genus, which was originally grouped in the Pseudomonas genus (^{Yabuuchi et al., 1992}). In the case of broccoli inflorescences, the tissue turned aqueous and with a bad smell, and the tissue displayed maceration three days after inoculation (Figure 4F), similar to the symptoms caused by Pseudomonas flourescens in this crop (^{Li et al., 2009}). In the case of ginger, no symptoms were observed, which shows an incompatible relationship.

Inoculation of bean with B. gladioli in a greenhouse. Necrotic lesions were observed in bean leaves and stalks (Figure 5). An incidence of 100% was registered, although no significant statistical differences were registered regarding the control in the number of flower buttons and height (Table 3). Recent studies indicate that some strains of the Burkholderia genus live endophytically in wild maize without causing any disease (^{Johnston and Raizada, 2011}). ^{Compant et al. (2008}) mention that in natural and artificial ecosystems, Burkholderia spp. may colonize the phylosphere and internal tissues of the plant, being the only bacterium that easily adapts to the environment and to plants. On the other hand, B. gladioli produces toxoflavin, which is highly allergenic to plants, fungi and animals (^{Anwar et al., 2017}). Likweise, ^{Lee et al. (2016}) indicate that the quorum sensing is related to the production of toxoflavin.

Burkholderia gladioli contains a genome made up of 9.05 Mb, two chromosomes and four plasmids, suggesting that the size of its genome is related to its ability to colonize different ecological niches (^{Mannaa et al., 2018}). In addition, Burkholderia sensu lato has different virulence factors such as type I, II, III, IV and V secretion systems, adhesins, pilis, siderophores, extracellular proteases, various polysaccharides, and quorum sensing molecules (^{Ferreira et al., 2019}).

Since B. gladioli survive in water, and most residual waters are used with agricultural purposes, this could lead to negative ecological consequences, such as a source of long-distance dissemination (^{Escobedo and Pardo, 2017}). When reisolating in the plant tissues studied in this investigation, in all cases, motility was found under the light microscope at 40X, coinciding with descriptions by ^{Chung et al. (2003}). This is an important characteristic in the pathogenicity of this bacteria, since it has polar flagella, which provides it with the advantage to move around the tissue of its host.

Figure 5 Symptoms induced by Burkholderia gladioli (MT672591) in bean plants (Phaseolus vulgaris). A) Start of the lesion; B) Progress of the lesion. 

In vitro sensitivity to chemical and biological products. None of the products tested had an inhibiting effect on B. gladioli (MT672591). This bacterium is characterized by its natural resistance to antibiotics, due to its ability to form biofilms and change its cellular cover to reduce membrane permeability, thus stopping the antibiotics from entering (^{Sousa et al., 2011}; ^{Torbeck et al., 2011}). The mechanisms that give it resistance include the production of ß-lactanases and other enzymes, as well as the modification of the target points of antibiotics (^{Ranjan et al., 2017}).

CONCLUSIONS

Burkholderia gladioli (MT672591) was identified morphologically, physiologically, biochemically, and molecularly as the causal agent of soft rot of the bulb of white onion.

B. gladioli colonized and induced damages in red onion bulbs, cactus cladodes, mushroom sporocarps, leek stalks, carrot roots and ginger rhizomes under laboratory conditions. Likewise, it promoted lesions in bean plants under greenhouse conditions. It also showed a natural resistance to agricultural cuprimicin, serenade powder, bactrol, agricultural bactriomicin, biological fungifree, cuprimicin 500, phyton, copper oxychloride, kasumin, final bacter, quatz and biological bacterbest. This study reports the polyphasic characterization of B. gladioli for the first time in Mexico. This bacterial species has the ability to grow at pH 4.0 and at 42 °C, which may give it a competitive ability in acid soils and semiarid conditions. It proved to have a wide range of hosts in postharvest conditions, as well as being resistant to some products used in the field for the control of phytopathogens.