Introduction

Rice (Oryza sativa) is one of the most important crops worldwide, since it is a staple food for more than 3.5 billion people (^{Faizal-Azizi and Lau, 2022}). The main countries that produce this crop are Continental China, India, Indonesia, Vietnam and Thailand (FAOSTAT, 2023). In Mexico, rice is one of the four main staple crops in the diet of the country’s population. Nationwide, approximately 40 thousand ha are planted, with more than 17 varieties of rice in diverse agroecological conditions, particularly in the states of Campeche, Jalisco, Morelos, Nayarit and Tamaulipas, with a production of 246 thousand t, strengthening the agro-industrial economy of the country (SIAP, 2023).

Worldwide, the rice crop is affected by different pathogens, although, Acidovorax avenae subsp. avenae, Burkholderia glumae and Xanthomonas oryzae pv. oryzae are among the main bacterial pathogens transmitted via seeds (^{Kang et al., 2016}). However,

B. glumae, the causal agent of the bacterial panicle blight (BPB), stands out for its worldwide distribution and economic importance (^{Cui et al., 2016}).

B. glumae has been reported in more than 18 countries as a highly destructive pathogen in rice (^{Zhou, 2019}) fields. In the last decade, BPB has been considered an emerging disease that has increased its incidence and severity in rice-producing countries due to climate change, leading to losses of up to 75% in the yield and quality of the grain in tropical and subtropical climates, particularly in regions with high temperatures and relative humidity (^{Abhishekh-Gowda et al., 2022}; ^{Mizobuchi et al., 2018}). In the Americas, the presence of B. glumae has been reported in Colombia, United States, Panama, Venezuela, Nicaragua, Ecuador, Costa Rica, the Dominican Republic and Puerto Rico (CABI, 2021).

B. glumae can survive in seeds, prosper in plant tissues, colonize different microenvironments in the plant and cause severe epidemics in rice crops, due to its high transmission efficiency and dissemination via seeds (^{Ortega and Rojas, 2021}). Consequently, the BPB disease begins with the infected seed as a primary inoculum. Subsequently, the pathogen is disseminated with the splashing of irrigation water and physical contact between healthy and diseased panicles (^{Ham et al., 2011}; ^{Hasegawa, 2012}; ^{Zhou, 2019}). On a global scale, the most frequently reported BPB symptoms are the glume blotch, the scarcity or absence of grain development, seed rot, abnormal root development and the rotting of seedlings (^{Sayler et al., 2006}), as well as blight in seedlings and panicles (^{Faizal-Azizi y Lau, 2022}; ^{Zhou-qi et al., 2016}).

Scientists of the National Forestry, Agriculture and Livestock Research Institute (INIFAP), in the Zacatepec, Morelos, experimental field are leaders in the genetic breeding of rice against diverse biotic and abiotic pathogenic agents. Since 2016, in the INIFAP field experiments in Zacatepec, Morelos, seeds of different rice varieties have been observed with glume blotches. This investigation assumes that these symptoms are caused by Burkholderia glumae, due to the similarity with glume blotches in rice seeds reported for this pathogen as a causal agent of the BPB disease (^{Ramachandran et al., 2021}).

The aim of this investigation was to characterize and identify the causal agent of the symptoms of glume blotch in rice varieties planted in Mexico. Identifying the causal agent of these symptoms in glumes in rice seeds is crucial to develop future prevention, management and genetic breeding strategies for rice cultivation in Mexico.

Materials and methods

Plant material. In 2022, 37 seed samples were analyzed with different percentages of blotch damage and necrosis in glumes in 11 varieties of rice planted in Mexico (Table 1, Figure 1), provided by the INIFAP- Zacatepec, Morelos, experimental field.

Figure 1 Symptoms of glume blotch in panicles and seeds in the rice varieties: A. MILAGRO FILIPINO, B. MORELOS A- 98, C. PACÍFICO FL-15, D. IRGA. Symptoms of blotch in seeds of the rice varieties: E. MILAGRO FILIPINO, F. MORELOS A-98.  

Isolation of bacteria. One gram of seed from every sample was disinfested with sodium hypochlorite (0.5% v/v) for 5 min and rinsed three times with sterile distilled water. The seeds were placed in beakers with 5 mL of PBS (Phosphate-Buffered Saline) (pH 7.2) and stirred at 125 rpm (ORBIT 1900, Labnet) for 1 h. Subsequently, they were macerated with a sterile mortar with 5 mL of PBS buffer and serial dilutions were carried out until 10^-2. From every dilution, 20 µL were spread using a Digralsky loop onto Petri dishes containing a sucrose-phosphate-glutamate (S-PG) semi-selective medium for B. glumae (^{Quesada-González and García-Santamaría, 2014}). The dishes were incubated at 30 °C for 48 h. From the bacterial growth, individual colonies with a phenotype similar to B. glumae (circular, convex colonies with smooth edges and a red-purple center) (^{Abhishekh-Gowda et al., 2022}) were transferred to King’s B and Wilbrinks culture media (^{Koike, 1965}) and incubated at 30 °C for 48 h. The colonies with a phenotype similar to B. glumae were preserved in cryogenic culture (-80 °C) in nutrient broth and glycerol (40 %) for further studies.

Physiological and biochemical characterization. The bacterial isolates were physiologically and biochemically characterized by API 20E (Biomerieux, Durhan, NC, U.S.A.), following the manufacturer’s instructions. Likewise, its pectinolytic activity in potato (Solanum tuberosum) was evaluated, along with the hypersensitivity (RH) in tobacco (Nicotiana tabacum cv. xanthi) (^{Goszczynska et al., 2000}).

Molecular identification. The genomic DNA of the bacterial isolations was obtained with the method CTAB 2% (^{Doyle and Doyle, 1990}) from pure cultures in King’s B medium with 48 h of growth at 28 °C, with modifications: During the lysis process, 300 µL of 2% CTAB a (1.4 M NaCl, 0.02 M EDTA, 0.1 M Tris-HCl pH 8, and 2 g of CTAB for every 100 mL of buffer) and 3 µL of Beta Mercaptoethanol at 10 % were added, and incubated at 55 °C for 30 min. In the precipitation, 200 µL of the aqueous phase were collected and 300 µL of isopropanol (100%) were added, at 4 °C, along with 30 µL of ammonium acetate (7.5 M), and gently mixed by inversion. The concentration and purity of the DNA samples were analyzed by UV light spectrophotometry at 260 nm with a NanoDrop^TM 2000 (Thermo Fisher Scientific). The integrity was evaluated by electrophoresis in agarose gel (Invitrogen^TM) at 1% at 70V for 90 min. Molecular identification was performed by endpoint PCR with 100 ng μL^-1 of DNA from each bacterial isolation using universal (8F-1492R) and specific (Bglu3F-BgluR, OgF-OgR) primers for B. glumae (Table 2). The PCR GOTAQ® Flexi DNA Polymerase (Ref. M8295, Promega) kit was used in a C1000 Touch TM Thermal Cycler (Eppendorf).

The PCR conditions for the 8F-1492R primers were an initial denaturalization at 94 °C for 5 min, followed by 35 cycles at 94 °C for 60 s, 57 °C for 45 s, 72 °C for 60 s and a final extension at 72 °C for 8 min. For the specific primers Bglu3F-Bglu3R, the PCR conditions were an initial denaturalization at 95 °C for 2 min, followed by 35 cycles at 95 °C for 30 s, 54 °C for 30 s, 72 °C for 30 s and a final extension at 72 °C for 7 min. For the primers OgF-OgR, an initial denaturalization at 95 °C for 2 min, followed by 35 cycles at 95 °C for 20 s, 63 °C for 30 s, 72 °C for 30 s and an extension at 72 °C for 7 min.

The amplified products were detected in a tris-ethylene-diaminetetraacetic acid (EDTA) agarose gel at al 1.5%, purified and sequenced bidirectionally with the three pairs of primers in the company Macrogen Inc. (Seoul, South Korea). The submitted sequences were analyzed and edited with the program BioEdit to generate the consensus sequences of each strain. The consensus sequences were compared with the nucleotide sequences in the National Center for Biotechnology Information (NCBI; https://www.ncbi.nlm.nih.gov/) database using the BLASTn 2.2 (Basic Local Alignment Search Tool) program. The polymorphism of the strains in the 16S rRNA was determined with the multiple alignment of the sequences from Mexico and those available in the NCBI with the program CLUSTAL W, implemented in the program MEGA 11 (Tamura et al., 2021). Based on the 16S rRNA sequences of the 11 strains, a phylogenetic tree was generated and analyzed with the Maximum Likelihood (ML) method 500 Bootstrap replications. The trees were generated using the sequence of a species of the Burkholderia genus as outgroup.

Pathogenicity in rice seed. A pure culture of the B. glumae CPM01 strain was used, identified in this study, with 48 h of growth in King’s B. The pathogenicity tests were carried out in Milagro Filipino variety seeds without any symptoms of glume blotch. The inoculation was carried out following the protocols by ^{Li et al. (2016}) and ^{Ura et al. (2006}) with some modifications. Fifteen seeds were disinfested with 70% ethanol (v/v), followed by 0.5% sodium hypochlorite (v/v) for 5 min and three washes with sterile distilled water. Subsequently, the seeds were inoculated by submersion in a bacterial suspension with 3.5 x 10⁸ CFU mL^-1 for 1 h and placed in germination boxes (BOX LARGE, THOMAS SCIENTIFIC, USA). The boxes were kept under laboratory conditions at a constant temperature of 30 °C and a relative humidity of >70%. Fifteen days after inoculation (dai), symptoms were found on the seed, root and stem of seedlings. The assay was carried out in a completely randomized design with three repetitions per treatment. Treatment 1 consisted of seeds inoculated with B. glumae strain CPM01 and the treatment of 2 non-inoculated seeds (submersion of seeds in sterile distilled water) (control).

Pathogenicity in rice seedlings. The pathogenicity tests were performed in healthy seedlings of the Milagro Filipino and INIFLAR-R varieties, which are extensively planted in Mexico (SIAP, 2023). The seedlings were obtained from healthy and disinfested seeds as described earlier. The seeds were planted in Styrofoam cups (1 kg) with sterile soil (NUTRIGARDEN). Inoculation was performed when the seedlings developed their fourth leaf (^{Flórez-Zapata and Uribe-Vélez, 2011}). Using a sterilized syringe, a lesion was made on the stem of the seedling and 3 mL of a bacterial suspension were inoculated by spraying, along with 3.5 x 10⁸ CFU mL^-1 of B. glumae (strain CPM01). The seedlings were maintained under greenhouse conditions at 28-30 °C, with a relative humidity

>70%. Fifteen days after inoculation, symptoms were observed on the stem. The assay was carried out in a completely randomized design with three repetitions per treatment. Treatment 1 consisted of Milagro Filipino seedlings inoculated with B. glumae strain CPM01. Treatment 2 consisted of INIFLAR-R seedlings inoculated with B. glumae strain CPM01; treatments 3 and 4, of non-inoculated Milagro Filipino and INIFLAR-R seedlings (with lesion on the stem and sprayed with 3 mL of sterile distilled water) (control). Out of the tissues (seed and stem) in which the symptoms developed, the inoculated bacteria was reisolated in King’s B culture medium and identified by PCR with the specific primers (Bglu3 and Og) for B. glumae.

Results

Isolation of B. glumae. In nine (24.3%) of the 37 rice seed samples analyzed, 11 bacterial colonies were isolated. In the sucarose-phosphate-glutamate (S-PG), the colonies displayed slow growth, a convex circular shape, with smooth edges and a red-purple center. These phenotypic characteristics are correlated with what has been reported in this culture medium for B. glumae colonies (^{Ura et al., 2006}; ^{Quesada-González and García- Santamaría, 2014}). In the King’s B and Wilbrinks culture media, the bacteria formed yellow colonies, raised circular colonies, mucoid and with smooth edges (Figure 2). In the three media, a yellow diffusible pigmentation was observed around the bacterial colonies.

Figure 2 Morphological characteristics of the bacterial colonies isolated from rice seeds with symptoms of glume blotch. Culture of the colonies at 28 °C for 72 h in a culture medium. A. Sucrose-phosphate-glutamate (S-PG), B. King’s B, C. Wilbrinks. 

Out of the 11 rice varieties analyzed, the bacterial strains with phenotypes similar to B. glumae were isolated from seven varieties (63.6%): GOLFO FL-16, INIFLAR R, INIFLAR RT, IRGA, Milagro Filipino, Morelos A-98 and PACÍFICO FL-15. No strains were isolated with this phenotype from four varieties: Azteca, Azteca 84, Pacífico B1 and Sauta (Table 3).

Physiological and biochemical characterization. The 11 strains were Gram-negative, they caused soft rot in potato and induced RH in tobacco. The results of the characterization by API 20E differentiated two groups based on their metabolic profiles. Group 1 differentiated strains (n= 8) isolated from the varieties GOLFO FL-16 (CPM04), INIFLAR R (CPM01), INIFLAR-RT (CPM02 and CPM03), Milagro Filipino (CPN01), Morelos A-98 (CPN02) and PACÍFICO FL-15 (CPM05 and CPM06). Group 2

differentiated the strains (n= 3) isolated from the varieteis IRGA (CPJ01) and Milagro Filipino (CPN01 and CPJ02) (Tables 3 and 4).

Molecular identification of B. glumae. From the DNA of the 11 isolated bacterial strains found, fragments of 1500, 402 and 174 bp were amplified with the 8F-1492R universal primers, and the specific primers for B. glumae Bglu3F-Bglu3R and OgF-OgR, respectively (Figure 3).

Figure 3 Detection of B. glumae by endpoint PCR. Partial amplification of the following genes: A. 16S rRNA with primers 8F-1492R (1500 bp). B. Putative protein with primers Bglu3 (174 bp). C. 16S-23S rRNA intergenic spacer region with primers Og (402 bp). Lane 1: INIFLAR-R (CPM01); lanes 2-3: INIFLAR-RT (CPM02 and CPM03, respectively); Lane 4: GOLFO FL-16 (CPM04); lanes 5-6: PACÍFICO FL-15 (CPM05 and CPM06, respectively); Lanes 7-9: Milagro Filipino (CPN01, CPN03 and CPJ02, respectively); Lane 10: IRGA (CPJ01) and lane 11: Morelos A-98 (CPN02). MM: Molecular marker 100 bp (GOLDBIO); (-): Negative control (sterile distilled water). 

The bioinformatic analysis of the 16S rRNA gene fragments, putative protein and the 16S-23S rRNA intergenic spacer region of the 11 strains indicated an identity value between 99 and 100% with B. glumae sequence CPO23203.1. reported in the NCBI GenBank (Table 5, Figure 4), which corresponds to the complete sequence of the genome of B. glumae strain 336gr-1, isolated from O. sativa in Louisiana, United States.

Figure 4 Phylogenetic analysis of the fragment of region 16S rRNA of 11 strains of B. glumae isolated in Mexico. Consensus phylogenetic tree inferred by the Maximum likelihood method. Model GTR+G+I and 500 Bootstrap replicates. The Burkholderia gladioli sequence CP098321.1 was included as an outgroup. ?= B. glumae strains isolated from seeds with symptoms of glume blotch from rice varieties planted in Mexico. 

The 16S rRNA nucleotide sequences of the 11 B. glumae strains were registered in the NCBI GenBank with accession numbers CPM01 (PV435843), CPM02 (PV435844), CPM03 (PV435845), CPM04 (PV435846), CPM05 (PV435847), CPM06 (PV435848), CPN01 (PV435849), CPN02 (PV435850), CPN03 (PV435851), CPJ01 (PV435852) and CPJ02 (PV435853).

Pathogenicity of B. glumae in rice seeds and seedlings. At 15 ddi, B. glumae CPM01 caused necrosis and rot in 100% of seeds, roots and stem of Milagro Filipino variety seedlings. Likewise, the bacteria inoculated in the Milagro Filipino and INIFLAR-R seedling stems caused necrosis and stem rot, as well as the abnormal development of roots and a reduction in seedling growth. No symptoms were observed in the seeds and seedlings of the control treatment (Figure 5).

Figure 5 Symptoms induced by B. glumae CPM01 15 dai in rice seeds and seedlings. MILAGRO FILIPINO variety: A. Control, submersion of seed in sterile distilled water. B. Necrosis and rotting of seeds, roots and stem. C. Reduction in growth and abnormal root development. INIFLAR-R variety: D. Control, spraying of sterile distilled water on stem. E. Necrosis and rotting of seedling stems. F. Reduction in growth and abnormal development of roots. 

The bacteria inoculated in seeds and seedlings were reisolated and identified by PCR as B. glumae with specific primers Bglu3 and Og described earlier, confirming the pathogenicity of the bacteria.

Discussion

Rice is a staple crop for over 50% of the world’s population. In this investigation, B. glumae was isolated and identified in rice seeds with symptoms of glume blotch. Out of 37 seed samples analyzed, bacterial colonies were isolated in nine of them (24.3%). The bacteria developed round, convex, smooth and yellow colonies with red-purple centers in the semiselective medium S-PG, which coincides with the morphology described for B. glumae in other investigations (^{Quesada-González and García-Santamaría, 2014}; ^{Ramachandran et al., 2021}; ^{Tsushima et al., 1986}). Likewise, all colonies produced a diffusible yellow pigment in the King’s B and Wilbrinks culture media. This pigment, known as toxoflavin, is a toxin that is considered a significant virulence factor for the pathogenicity of B. glumae in rice plants (^{Gil et al., 2022}; ^{Maurya et al., 2023}).

The identification of B. glumae was confirmed by PCR with universal and specific primers. Based on the partial sequences of the 16S rRNA genes, putative protein and 16S- 23S rRNA intergenic spacer region, the 11 isolated strains displayed a coverage of 100% and identity with the B. glumae strain CP023203.1 deposited in the NCBI. The analysis of specific sequences for these bacteria, along with the putative protein gene (Bglu3), enables the precise identification of B. glumae and its discrimination among B. gladioli, which present a phenotypical similarity with the B. glumae colonies, as well as in the mechanism of biosynthesis of toxoflavin, causing identical symptoms to those of B. glumae in rice seeds and panicles (^{Lee et al., 2018}; ^{Nandakumar et al., 2009}). Likewise, the analysis of the sequence of the 16S-23S rRNA intergenic spacer region (Og) helps separate B. glumae from other pathogens transmitted by rice seeds, such as Xanthomonas oryzae pv. oryzae and Acidovorax avenae subsp. avenae (^{Kang et al., 2016}). Burkholderia glumae is a pathogen that is transmitted through seeds (^{Cui et al., 2016}; ^{Chompa et al., 2022}) and it has been proven that the infection of rice seeds is the main source of dissemination of the bacteria within a country (^{Abhishekh-Gowda et al., 2022}). The presence of B. glumae is estimated to cause losses of up to 75% in grain yield and quality, depending on the rice variety in regions with high temperatures and relative humidity (^{Fory et al., 2014}; ^{Shew et al., 2019}), as in the case of the regions in Mexico where rice is planted.

In this investigation, B. glumae was isolated from seeds from seven (63.6%) rice varieties: GOLFO FL-16, INIFLAR-R, INIFLAR-RT, IRGA, Milagro Filipino, Morelos A-98 and PACÍFICO FL-15. The bacteria was isolated with a higher frequency (43%) from the seeds of the variety Milagro Filipino, which is highly relevant, since this variety currently covers 86.8% of the surface planted in Mexico and is cultivated in eight of the 12 rice-producing states, followed by the Morelos variety, with 7.6% of the surface planted nationwide (SIAP, 2023).

So far, neither the chemical and biological control methods nor cultural practices have been efficient for the optimum management of BPB (^{Mizobuchi et al., 2018}; ^{Ortega and Rojas, 2021}). The genetic resistance of plants is considered one of the main strategies for the control of bacterial diseases; however, on a global scale, no rice varieties or hybrids have been found that are completely resistant to B. glumae. Several varieties have been evaluated in order to identify materials with some degree of tolerance or resistance to the bacteria, although the varieties planted have shown high levels of susceptibility and moderate levels of resistance (Mizobuchi et al., 2018; ^{Shew et al., 2019}). Nevertheless, resistance to B. glumae in rice has been reported to be quantitative (associated to multiple genes), and therefore the mapping of Quantitative Trait Locus (QTL) has been suggested in rice varieties with these levels of resistance (^{Pilet-Nayel et al., 2017}).

The inoculation of B. glumae strain CPM01 identified in this study confirmed the pathogenicity of the bacteria in rice seeds and seedlings of the Milagro Filipino and INIFLAR- R varieties, both of which have been planted extensively in Mexico. A greater severity was observed in seeds, roots and stems of the Milagro Filipino variety, which suggests that there could be different responses of susceptibility to the infection of B. glumae in seedlings of the rice varieties planted in Mexico.

To date, there are no reports of B. glumae causing the BPB disease in rice in Mexico. However, this study shows the presence of B. glumae in seeds with symptoms of glume blotch in seven of the 11 rice varieties planted in Mexico, including the variety Milagro Filipino, which covers 86.8% of the national surface planted (SIAP, 2023). Due to this, it is necessary to further investigate the origin and route of entry into the country, as well as the genetic variability and virulence of the B. glumae populations identified in this study. In countries with high rice production rates, such as Thailand, Malaysia and India, knowledge on the genetic variability and virulence of B. glumae has been important to develop strategies for the management of BPB (^{Jungkhun et al., 2021}; ^{Kumar et al., 2023}; ^{Ramachandran et al., 2021}; ^{Sreenayana et al., 2024}).

In 2016, rice genetic materials were evaluated under conditions of the dry and humid Mexican tropics. From this evaluation, the varieties INIFLAR- RT, INIFLAR-R, Pacífico FL-15 and Golfo FL-16 were selected, as they displayed a high potential for yield, a high industrial quality of the grain and a moderate resistance to Magnaporthe oryzae (synonymous with Pyricularia oryzae), to the rice hoja blanca virus (RHBV) and to Bipolaris oryzae, causal agents of the “rice blast”, “hoja blanca” y “brown spot” diseases, respectively (INIFAP-Fondo Sectorial SAGARPA-CONACyT, 2016). Nevertheless, this evaluation did not consider B. glumae.

The incorporation of resistance to B. glumae in the rice varieties requires a careful evaluation of the response of the available genetic materials against the pathogen, in order to identify those with the highest resistance (^{Zhou, 2019}). Due to the global relevance of

B. glumae in rice, in Mexico it is crucial to evaluate the level of resistance against B. glumae of the rice varieties currently planted, since the degree of susceptibility to this pathogen under the environmental conditions that prevail in the areas in which rice is grown is unknown. As a consequence, the results of this investigation provide relevant information for the production of rice in Mexico, since the early and accurate detection of B. glumae in seeds is considered a crucial prerequisite for the development of epidemiological studies, as well as to limit its dissemination and to implement strategies to contain, prevent and for the genetic breeding of rice varieties with resistance to this pathogen (^{Abhishekh-Gowda et al., 2022}; Faizal-Azzizi and Lau, 2022; ^{Kumar et al., 2023}).

Conclusions

Burkholderia glumae is the causal agent of glume blotch in rice seeds and was found in seven of the eleven rice varieties planted in Mexico (GOLFO FL-16, INIFLAR-R, INIFLAR-RT, IRGA, Milagro Filipino, Morelos A-98 and PACÍFICO FL-15). The bacterium was more prevalent in the variety Milagro Filipino, which is one of the most extensively planted varieties in different regions of the country. This investigation is the first to report the presence of B. glumae in Mexico in rice seeds with symptoms of glume blotch. These results are of great importance for the development of future investigations centered on the establishment of containment and management strategies, as well as on the genetic breeding of rice varieties planted in Mexico with resistance to B. glumae.