Prevalence of Burkholderia glumae in rice crops in Ecuador 1,2

Burkholderia glumae is the agent responsible for bacterial panicle blight disease (BPBD) of rice that causes severe damage to this crop worldwide. During 2012 and 2013, symptoms of BPBD were observed in Palestina city, located in Guayas province, Ecuador. In 2014, the presence of B. glumae was confirmed at this location. In view of the socio-economic importance of rice in Ecuador, this research aimed to investigate the prevalence of B. glumae in other rice-producing regions. Eighteen bacterial isolates obtained from blighted kernels were characterized. Physiological, biochemical, serological, and molecular assays and the amplification of the 16S-23S rRNA ITS of the bacterial isolates collected confirmed the identity of the BPBD-associated bacterium. Pathogenicity assays verified the ability of these isolates to produce discoloration, spotting, and empty grains, symptoms associated with BPBD. Antibiotic assays showed that EC-EELS-01 isolate was sensitive to ciprofloxacin and tetracycline, and resistant to polymyxin. The dissemination and prevalence of B. glumae were confirmed in the rice- producing areas of El Oro, Cañar, Guayas, and Los Rios provinces. This research will serve to develop genetic studies to characterize the population in the B. glumae isolates in Ecuador.


INTRODUCTION
Bacterial panicle blight disease (BPBD), also called bacterial grain rot or panicle blight, is considered a major disease of rice (Oryza sativa L.), causing severe damage to this crop worldwide (Nandakumar et al., 2009;Lee et al., 2016a;Bigirimana et al., 2015;Mondal et al., 2015). Bacterial panicle blight disease was first reported in Japan, causing grain rotting and seedling blight on rice (Goto and Ohata, 1956). Since then, the disease has been reported in several countries of Africa, Asia, South and Central America; shifting from being considered a minor plant disease to a major problem due to environmental conditions (Cui et al., 2016b). The causal agent of this disease can be either a single Burkholderia species or a combination of several, which are mainly seed-transmitted. At present, there are no chemical options to control bacterial infected rice crops (Ham et al., 2011;Lee et al., 2016a).
Burkholderia glumae is the major causal agent of BPBD on rice, belonging to the Class II: Betaproteobacteria, Order Burkholderiales of the Family Burkholderiaceae (Kurita and Tabei, 1967). Rice is considered its main host plant (Lee et al., 2016a;Magbanua et al., 2014;Sharma et al., 2013). It can survive on rice leaves and sheaths, spreading upwards as the plant grows (Ham et al., 2011). Seedling rot induction, grain discoloration, grain rot and leaf-sheath browning, as well as flower sterility and decrease of grain weight are typical symptoms of the disease (Hikichi, 1993;Cottyn et al., 1996). Daytime temperatures > 32° C, nighttime temperatures > 25° C, and high humidity contribute to disease development (Cui et al., 2016). Apart from the detrimental effects on rice crops, this bacterium is also responsible for wilting of other crops, such as Capsicum annuum, Sesamum indicum, Solanum lycopersicum, S. melongena and Perilla frutescens var. japonica Hara (Jeong et al., 2003).
Burkholderia glumae is an aerobic motile, gram-negative bacterium that possesses two to four polar flagella, generally of 0.5-0.7 x 1.5-2.5 μm in diameter. The production of toxoflavin and lipases, along with the polar flagella represent important virulence determinants (Jeong et al., 2003;Nandakumar et al., 2009;Lee et al., 2016b). On potato agar, the bacterium can produce a fluorescent pigment. Bacterial growth temperature ranged between 11 and 40° C, having an optimal temperature between 30 and 35° C. It is able to hydrolyze gelatin, while nitrate reduction, arginine dihydrolase, starch hydrolysis, and hydrogen sulfide production are negative (Palleroni, 2015;Urakami et al., 1994). Acids are produced from various sources, such as arabinose, glucose, fructose, galactose, mannose, xylose, glycerol, mannitol, and inositol, while no acid is produced from rhamnose, sucrose, maltose, lactose, raffinose, dextrin, starch, inulin or salicin (Palleroni, 2015). On YPDA medium, colonies produced a yellow pigment soluble in chloroform (Ura et al., 2006). Burkholderia glumae usually grows at 40° C on a culture medium containing 3% NaCl and utilizes several compounds as carbon source [for a complete list, please, review Palleroni (2015)]. Colonies do not produce fluorescent pigment on King B medium. Nevertheless, an intense yellow pigment is produced, which diffuses through the media due to the production of toxoflavin (Nandakumar et al., 2009;Lee et al., 2016b). Rice spikelets infected by this phytotoxin display symptoms of brown stripes on both the palea and lemma, which are modified stems that protect flower organs, inhibiting the growth of whole rice plants (Iiyama et al., 1995).
In 2013, typical symptoms of stained, discolored and abortive or unfilled grains were observed in rice panicles grown at Palestina city of Guayas province of Ecuador. Riera et al. (2014) reported B. glumae as the disease causal agent resulting in yield and grain quality losses. Burkholderia glumae has also been reported in other countries, such as Colombia (Zeigler and Alvarez, 1989), Panama (Nandakumar and Rush, 2007), USA (Nandakumar et al., 2009), Dominican Republic, Venezuela, Brazil, Nicaragua and Costa Rica (Cui et al., 2016). Other pathogens cause symptoms similar to those of B. glumae. Pseudomonas fuscovaginae, for example, also causes staining of grains and sheaths, and even destruction of the flower reproductive organs (Zeigler and Alvarez, 1987a). Furthermore, P. fuscovaginae, like B. glumae, also causes sheath rot and grain discoloration, which can make it difficult to distinguish symptoms and damages caused by these two species (Cottyn et al., 1996). The primary goal of this research was to study the prevalence of B. glumae in rice fields in Ecuador and to characterize 18 isolates causing BPBD.

Isolation of bacterial isolates from symptomatic rice seeds
From 2014 to 2015 rice panicles presenting discolored and unfilled grain symptoms (Figures 1A to C) were harvested at four localities: Cañar, El Oro, Guayas and Los Ríos, Ecuador (Table 1). From each plant sample, 10 symptomatic rice grains were selected, placed in 1.5 mL microtubes containing 1 mL sterile water and incubated for 4 h at 37° C. After incubation, an aliquot of these samples was streaked on King B medium (King et al., 1954) and incubated at 37° C. Bacterial colonies showing circular shape, cream color, smooth and convex elevation, with translucent and complete borders were selected and isolated in pure culture. Gram-negative test was performed by a solubility assay in 3% KOH (Suslow et al., 1982). Selected colonies were grown on bacteria screening medium 523 (Kado and Heskett, 1970).

Serological and molecular analysis
Indirect-ELISA (enzyme-linked immunosorbent assay) using a specific antibody against B. glumae (Agdia, Inc.) and PCR assays using specific primers were performed. A single colony was grown in Luria-Bertani (LB) liquid medium for 24 h at 37º C to isolate total DNA using Wizard® Genomic DNA Purification Kit 6 (Promega). PCR analysis was carried out in an Agilent Technologies Sure Cycler 8800 thermal cycler. Amplification of the 16-23S rRNA ITS region of the B. glumae isolates was performed using specific primers to the Burkholderia species (forward) 5'-ACG TTC AGG GAT RCT GAG CAG-3' and (reverse) 5'-AGT CTG TCT CGC TCT CCC GA-3' (Sayler et al., 2006).
The PCR mix was composed of 1X GoTaq Flexi Buffer, 1.5 mM MgCl 2 , 0.2 μM of dNTPs (Promega), 0.16 μM of each primer (forward and reverse), 0.5 U GoTaq® Flexi DNA Polymerase (Promega), and 10 ng of template genomic DNA. PCR products were separated by 1.2% agarose gel electrophoresis (Promega) and immersed in Diamond TM Nucleic Acid Dye intercalating solution (Promega). One Kb Plus DNA Ladder (Invitrogen) was used to estimate the length of PCR products.

Pathogenicity assays
Four bacterial isolates (i.e., EC-EELS-01, -02, -03, and -07) were randomly selected to carry out pathogenicity assays using rice variety INIAP 15 and to confirm Koch's postulates. A single colony was used to obtain bacterial growth onto 523 solid media, resulting colonies were re-suspended and homogenized in sterile 0.85% NaCl solution, and the OD 600 was adjusted to 0.2. Inoculations were performed on panicles in the anthesis stage using a manual sprayer, and the inoculated plants were immediately placed in a humid chamber at 42.8° C with 72.6% relative humidity. Control plants were mock inoculated with sterile water. After three days, typical symptoms were recorded and percentage of unfilled grains was assessed. Successful infection by B. glumae isolates was confirmed by PCR (as described previously) from inoculated panicle grains.
Statistical analysis was performed using a completely randomized design and mean data were analyzed using Tukey test (p-value > 0.05) by InfoStat software (Di Rienzo et al., 2016).

Isolation of the bacterial isolates and physiological, biochemical, nutritional characterization
Eighteen bacterial isolates were obtained from four localities sampled: El Oro, Cañar, Guayas and Los Ríos, Ecuador (Table 1). Colonies were cream in color, convex with a smooth surface and translucent border. All were considered Gram-negative bacteria (Table 2).
Colonies did not show enzymatic hydrolysis of fluorescein diacetate on King B medium. However, they displayed a diffusible yellowish and greenish color, which is typical of the toxoflavin toxin, confirmed by strong production of this pigment on YPDA medium. In addition, all isolates grown on NBY medium at 40° C and on YDC medium were light cream in color (Table 2). Non-bacterial growth was observed of isolates EC-EELS-06 and EC-EELS-10 in LB medium at pH 4.0, and isolates EC-EELS-10 and EC-EELS-12 at pH 8.0, while at pH 9.0 none of the 18 isolates grew. Positive aerobic growth, catalase reaction, gelatin liquefaction, and ability to metabolize D-xylose, D-trehalose, and D-sorbitol as carbon sources were confirmed for all isolates (Table 2). All isolates were negative to starch hydrolysis with the exception of isolate EC-EELS-07 that did not display pectinolytic activity on potato slices (Table 2).

Serological and molecular detection
ELISA and PCR results confirmed the identity of B. glumae ( Table  2). The PCR results for the 18 isolates using specifically designed primers (Sayler et al., 2006) for B. glumae showed an amplicon of expected size (~282 bp) ( Figure 1G).

Antibiotic susceptibility assays
Assays of sensitivity to antibiotics showed that EC-EELS-01 isolate is sensitive to ciprofloxacin and tetracycline causing inhibition halos of 37.7 and 40.5 mm, respectively. Furthermore, B. glumae isolate showed resistance to polymyxin B and a certain tolerance for ampicillin (9.0 mm), erythromycin (12.8 mm), trimethoprim (22.4 mm), and gentamicin (23.1 mm) ( Figure 1H).

DISCUSSION
Our results from physiological, biochemical, serological, and molecular assays showed that all 18 isolates had typical characteristics of B. glumae in agreement with reports by Palleroni (2015) and Schaad et al. (2001). The results confirmed the spread and prevalence of B. glumae, which has caused severe damage in the main rice producing areas of the provinces of El Oro, Cañar, Guayas and Los Ríos, Ecuador. Burkholderia glumae was first reported in Japan (Goto and Ohata, 1956) and subsequently detected in several other countries. In 2013 it was observed in the rice plantations of Palestina city, Guayas province, Ecuador (Riera et al., 2014). Since then, symptoms of BPBD on Ecuadorean rice plantations have been widespread and, most of the time, were associated with P. fuscovaginae (INIAP, 1987). In this research, we characterized 18 B. glumae isolates harvested in associated with BPBD.
Bacterial isolates growth at the different pH showed lack of, or poor growth at extreme pH conditions, which corresponds to established parameters of B. glumae identification (Urakami et al., 1994;Palleroni, 2015). All isolates on NBY medium supplemented with 1.5, 2.0 or 2.5% NaCl developed, while no growth was observed at 3%, showing lower tolerance to NaCl compared with isolates profiled in other studies that showed tolerance at 3% NaCl (Zhou, 2014).
Pathogenicity assays, using four isolates randomly selected (i.e., EC-EELS-01, -02, -03, and -07), confirmed their ability to cause typical symptoms such as discoloration, stained, and unfilled grains in rice panicles (up to 81.19%) of the rice variety INIAP 15 ( Figure 1D to F). These results are consistent with grain damage (93.8%) observed in the rice variety XXI (Fory et al., 2014). Mean temperature at Ecuadorian areas, where samples were collected, was 27.0° C (INAMHI, 2016). Under elevated temperature and humidity conditions, B. glumae causes gynoecium wilt and deformation, along with abortion of pollen grains because of colonization of the palea and lemma interior causing sterility (Li et al., 2017). Toxoflavin production causes grain rotting and discolored panicles (Ilyama et al., 1995;Jeong et al., 2003;Luo et al., 2007;Mondal et al., 2015;Lee et al., 2016b). In our analysis, toxoflavin production by B. glumae was indicated by the production of a yellow pigment on a King's B agar plate.
In agreement with Bauer et al. (1966), B. glumae showed sensibility to tetracycline (that blocks binding of aminoacyl-tRNA to A site on the ribosome) and ciprofloxacin (a potent inhibitor of DNA gyrase). Oxytetracycline has the ability to inhibit bacterial growth without killing it and has been used to treat broccoli and cabbage infested with Xanthomonas campestris reducing infection to less than 1% (Dekker, 1963). Similarly, treatment of rice seeds with oxolinic acid sprayed before and after panicle production has proved to be efficient in BPBD management (Hikichi, 1993). Nonetheless, some strains of B. glumae have been identified as being naturally resistant to oxolinic acid due to a gyrA mutation in the DNA gyrase (Naughton et al., 2016). However, Cui et al. (2014) demonstrated that metallic copper has an antibacterial effect against B. glumae by causing a buildup of copper in the bacterial cell, losing cell membrane integrity by directly affecting proteins and lipids, in addition to DNA degradation. Burkholderia glumae strain PG1 has specialized genes that encode degradation systems for toxic substances like atrazine, nitrotoluene, and nicotinate; as well as an efflux system for pumping out a broad range of antibiotics (Lee et al., 2016b). Therefore, it is necessary to explore genetic rice materials that show resistance to BPBD and investigate chemical molecules that can act as bactericides in disease management.
In summary, we identified 18 bacterial isolates of B. glumae, a pathogen associated with typical symptoms of BPBD. The bacterium is widespread in rice-producing areas in Ecuador. Symptoms observed in the field confirmed that B. glumae is one of the causal agents of vanishing panicles and decreased in rice yields. Currently, it is considered the second most important phytobacterium infecting rice in the country, following P. fuscovaginae.
Therefore, there is a compelling need to evaluate sources of rice germplasm resistance and use of pathogen-free seed. Thus, studies of genetic variability among B. glumae isolates are extremely important to determine the existence of population subdivision. Currently, our research team is sampling other rice regions aiming to develop a broad genetic study to establish population structure of B. glumae isolates in Ecuador.