Abstract
Background The genus Pantoea forms a complex of more than 25 species, among which several cause diseases of several crop plants, including rice. Notably, strains of Pantoea ananatis and Pantoea stewartii have been found to cause bacterial leaf blight of rice in Togo and Benin, while other authors have observed that Pantoea agglomerans can also cause bacterial leaf blight of rice. The contribution of these and perhaps other species of Pantoea to plant diseases and yield losses of crop plants is currently not well documented, partly due to the lack of efficient diagnostic tools.
Result Using 34 whole genome sequences of the three-major plant-pathogenic Pantoea species, a set of PCR primers that specifically detect each of the three species, P. agglomerans, P. ananatis, and P. stewartii, was designed. A multiplex PCR protocol which can distinguish these three species and also detects members of other Pantoea species was further developed. Upon validation on a set of reference strains, 609 suspected Pantoea strains that were isolated from rice leaves or seeds originating from 11 African countries were screened. In total, 41 P. agglomerans strains from eight countries, 79 P. ananatis strains from nine countries, 269 P. stewartii strains from nine countries and 220 unsolved Pantoea strains from ten countries were identified. The PCR protocol allowed detecting Pantoea bacteria grown in vitro, in planta and in rice seeds. The detection threshold was estimated at 5 ng/mL of total genomic DNA and 1 × 105 CFU/mL of heated cells.
Conclusion This new molecular diagnostic tool will help accurately diagnose major plant-pathogenic species of Pantoea. Due to its robustness, specificity, sensitivity, and cost efficiency it will be very useful for plant protection services and for the epidemiological surveillance of these important crop-threatening bacteria.
Background
The genus Pantoea was first described in 1989 and was recently taxonomically classified as a member of the Erwiniaceae family [1]. More than 25 species of this genus have been described and reported worldwide [2,3]. Etymologically, the genus name Pantoea is derived from the Greek word ‘Pantoios’, which means “of all sorts or sources” and reflects the diverse geographical and ecological sources from which the bacteria have been isolated. Several species of the genus are qualified as versatile and ubiquitous bacteria because they have been isolated from many different ecological niches and hosts [2,4]. Remarkably, some species have the ability to colonize and interact with members of both the plant and the animal Kingdom [5]. Among the plant-interacting species, Pantoea ananatis, Pantoea agglomerans and Pantoea stewartii are well known for their phytopathogenic characteristics. They are recognized as the causal agent of several diseases, such as leaf blight, spot disease, dieback, grain discoloration, seed stalk rot, center rot, stem necrosis, palea browning, bulb decay etc. and affect several economically important crops, including cereals, fruits and vegetables [2,6,7].
Bacterial leaf blight caused by Xanthomonas oryzae pv. oryzae is an important disease of rice and affects rice cultivation in most regions of the world were rice is grown. The bacterium has been associated with this disease since a very long time [8]. Surveys were conducted from 2010 to 2016 to estimate the extent and importance of the disease and the phytosanitary status of rice fields in West Africa. While leaves showing bacterial blight (BB)-like symptoms were frequent, isolation or molecular detection of xanthomonads using the Lang et al diagnostic tool [9] often failed. Instead, other bacteria forming yellow colonies were observed and turned out to belong to the species P. ananatis or P. stewartii, as documented for samples from Togo and Benin [10,11]. Additionally, other cases of BB and grain discoloration caused by Sphingomonas sp. and other undescribed species have been detected in several sub-Saharan Africa countries [12]. This situation represents an “emerging” bacterial species complex that may constitute a threat to rice production in Africa. Therefore, a robust, specific, sensitive, and cost efficient diagnostic tool is of primary importance for accurate pathogen detection. However, none of the several simplex and multiplex PCR tools [13–18] and other molecular [19–24], physiological, biochemical [24–28] diagnostic tools available for Pantoea allows accurate simultaneous detection of the three major plant-pathogenic Pantoea species. Some of these methods are poorly reproducible and often limited to a single species while others are reproducible but again limited to one species or are not suited to doubtlessly detect African strains.
To overcome this unsatisfying situation, a molecular method was set up for detecting in a single reaction the three major plant-pathogenic Pantoea species (P. ananatis, P. stewartii and P. agglomerans), as well as other members of the genus. A universal multiplex PCR tool was therefore developed and first tested in silico on available genome sequences and on a set of reference strains from USA, Brazil, Spain and Japan. Afterwards, 609 suspected Pantoea strains from eleven Africans countries were evaluated with the newly described diagnostic tool. P. agglomerans was detected in rice leaves from several African countries for the first time. Finally, the specificity and sensitivity of the multiplex PCR was monitored by analyzing serial dilutions of genomic DNA, serial dilutions of bacterial cell suspensions and solutions of ground leaves and seeds that had been artificially or naturally infected. This new diagnostic tool will prove useful for phytosanitary services in routine diagnostics of Pantoea spp in any type of sample (e.g. leaves, seeds, soil, water).
Materials and Methods
Bioinformatics prediction of specific PCR primers
Pantoea genome sequences were retrieved from NCBI GenBank (Table 1). Sequences for housekeeping genes were identified by TBLASTN [29]. Sequences were then aligned with MUSCLE [30] at EMBL-EBI [31]. Diagnostic primers that can differentiate the three species, P. agglomerans, P. ananatis and P. stewartii, and one primer pair that would amplify DNA from the whole Pantoea genus were designed manually. The Tm for PCR primers were automatically predicted by Tm calculator tool at http://www.thermoscientificbio.com/webtools/multipleprimer/ which was developed based on the modified nearest-neighbor interaction method [32].
Optimization of the multiplex PCR
Different types of samples including total genomic DNA, bacterial cells, symptomatic rice leaves, as well as discolored and apparently healthy rice seeds were analyzed. Plant material was ground and macerated before use. To develop a multiplex PCR scheme, individual primer pairs were first tested against the different samples mentioned above, using annealing temperatures close to the predicted Tm (Tm ± 5 °C) and with progressive number of PCR cycles (25 to 35). Primer pairs were then mixed from duplex to quintuplex and PCR conditions were evaluated, testing annealing temperatures close to the optimal Tm of the individual primer pairs (Tm ± 3 °C) and various numbers of PCR cycles. At the end, three promising combinations of annealing temperatures and numbers of PCR cycles were re-evaluated in simplex PCR with the samples mentioned above. The best combination with high specificity and without background amplification was selected as the new diagnostic tool (Tables 2 to 4).
Evaluation of the sensitivity of the multiplex PCR scheme using genomics DNA and heat cells
Simplex and multiplex PCR were then used to evaluate the sensitivity of all the species-specific primer pairs individually or in combination with the genus-specific and the 16 sRNA primer pairs. Serial dilutions of total genomic DNA and heated bacterial cells were used for this evaluation. Three Pantoea strains, P. ananatis strain ARC60, P. stewartii strain ARC222, and P. agglomerans strain CFBP 3615, were used and distilled sterilized served as a negative control.
To evaluate the PCR scheme on live plant material, leaves and seeds were artificially infected with strains of the three Pantoea species. Rice leaves of the cultivar Azucena were inoculated as described previously [10,11]. To produce contaminated seeds, early maturity panicles of the Azucena rice cultivar were spray-inoculated with a 5%-gelatinized bacterial solution (106 CFU/mL). Distillated and gelatinized (5%) sterile water served as a negative control. Three weeks post inoculation, approximately 40% of the grains in the panicles exhibited discolorations. Panicles inoculated with sterile distilled water showed no symptoms. A total of five grains whose surface was first treated with a solution of hypochlorite (10%) and ethanol (70%) and then rinsed with sterile distilled water were ground in 100 mL of sterile distilled water. After centrifugation, the supernatant was used for PCR.
Evaluation of the multiplex PCR scheme on a large collection of African Pantoea strains
Bacterial strains used in this study are listed in Additional file 1. In total, 615 Pantoea strains from eleven Africans countries (Benin, Burkina Faso, Burundi, Ghana, Ivory Coast, Mali, Niger, Nigeria, Senegal, Tanzania, Togo) and seven reference strains from USA, Brazil, Spain and Japan were analysed by the new diagnostic tool. The African strains were isolated from rice leaves with BB symptoms, and from discolored and apparently healthy rice seeds. The samples had been collected from 2008 to 2016 in the main rice-growing areas of the countries. Other bacteria, including Xanthomonas spp, Sphingomonas spp, Escherichia coli, Erwinia spp, Burkholderia spp, and Pseudomonas spp, were used as controls. The strains were purified as single colonies, individually grown and preserved as pure cultures following routine methods [33]. Bacterial colonies were grown for 24 to 48 h on PSA plates containing 10 g peptone, 10 g sucrose, 16 g agar and 1 g glutamic acid per liter. Total genomic DNA was extracted using the Wizard genomic DNA purification kit (Promega) according to the manufacturer’s instructions. DNA quality and quantity were evaluated by agarose gel electrophoresis and spectrophotometry (Nanodrop Technologies, Wilmington, DE).
Results
Development of a diagnostic PCR scheme for plant-associated Pantoea
We aimed at designing diagnostic PCR primers that would target conserved housekeeping genes. The rationale behind was that these genes should be present in all strains, including genetic lineages that have not yet been discovered and would not be present in any strain collection. At the same time, we knew from previous work that sequences of housekeeping genes are divergent enough to doubtlessly distinguish and identify Pantoea strains at the species level.
A diagnostic Pantoea multiplex PCR method was developed in two steps. First, a complete inventory of publicly available Pantoea genome sequences was compiled, consisting of nine P. agglomerans, 14 P. ananatis, and three P. stewartii sequences, totaling to 26 whole genome sequences (Table 1). Complete coding sequences of four housekeeping genes that have previously been used for multilocus sequence analyses (MLSA) of Pantoea species [2], atpD, gyrB, infB, and rpoB, were then extracted and aligned. Sequence regions that were conserved in all strains of one species but were significantly different in the other two species were identified manually and chosen to design PCR primers (Table 2). To allow multiplexing, we made sure that the amplicon sizes would be between 400 and 750 bp and different enough to be easily distinguishable from each other upon gel electrophoresis (Fig. 1). As a positive control for the PCR reaction, one primer pair was included that would amplify DNA from all bacteria belonging to the Pantoea genus, resulting in a smaller amplicon of less than 400 bp. Finally, as a second control, a primer pair was included that targets the ribosomal 16S rRNA gene and leads to an amplicon that is larger than the four Pantoea-specific amplicons.
In the second step, all primer pairs (Table 2) were evaluated, first by simplex PCR and then by multiplex PCR, with increasing number of primer pairs, as explained in Material and Methods. Three Pantoea reference strains were used to develop the PCR scheme using genomic DNA and heat-inactivated bacteria: P. agglomerans strain CFBP 3615, P. ananatis strain ARC60 and P. stewartii strain ARC222 (Fig. 2). Agarose gel electrophoresis demonstrated that the multiplex PCR was able to detect and distinguish all three Pantoea species. Notably, the multiplex PCR scheme was also able to detect two or three Pantoea species when the corresponding species were present in the same template DNA, as demonstrated by PCR reactions containing equal amounts of DNA of the different species (Fig. 2).
To simplify the analyses and to avoid isolation of bacteria from plant samples, thus reducing the costs per sample, the PCR scheme was also evaluated on infected leaf material and contaminated seeds. As shown in Fig. 3, the multiplex PCR was able to doubtlessly detect all three Pantoea species in both types of plant samples, as demonstrated for the strains CFBP 3615 (P. agglomerans), ARC60 (P. ananatis), and ARC222 (P. stewartii). At the end, a robust PCR protocol was available that was able to amplify DNA from total genomic DNA, bacterial cells, symptomatic rice leaves and from infected rice seeds.
Evaluation of the sensitivity of the multiplex PCR scheme using genomic DNA and heated cell suspensions
The evaluation by simplex and multiplex PCR showed that all the species-specific primers were very sensitive individually or in combination with the genus-specific and the 16 sRNA universal primers (Fig. 4). The most sensitive primer pair in simplex PCR was the one targeting P. stewartii with a detection limit of 5 pg under our experimental conditions, followed by the P. agglomerans-specific primer pair (detection limit of 50 pg) and the P. ananatis-specific primer pair (detection limit of 0.5 ng). A similar trend was observed in the multiplex PCR on genomic DNA, with the same detection limit as in simplex PCR for P. stewartii and P. ananatis and a tenfold less sensitivity for P. agglomerans.
When heated bacterial cell suspensions were used as template, the P. ananatis-specific primer pair was the most sensitive allowing detection of 103 CFU/mL, while the other two primer pairs were able to detect 104 CFU/mL. However, when all five primers pairs were used in multiplex, the sensitivity was very similar for all three species with a detection limit of approximately 104 CFU/mL.
Evaluation of the multiplex PCR scheme on a large collection of African Pantoea strains
Because recent surveys had indicated that Pantoea species could be responsible for many unsolved infections of rice fields in sub-Saharan Africa [10,11], we screened a large collection of isolates. We first re-evaluated a few African strains that had been identified as P. ananatis (ARC22, ARC60, ARC651) and P. stewartii (ARC229, ARC570, ARC646), using species-specific and the genus-specific PCR primers [10,11]. The multiplex PCR scheme confirmed their previous taxonomic classification. Next, we screened a large collection of African bacterial isolates from rice samples (>1000 strains) among which 609 strains were found to belong to the genus Pantoea (Additional file 1). Specifically, this work diagnosed 41 P. agglomerans strains from eight countries (Benin, Ghana, Mali, Niger, Nigeria, Senegal, Tanzania, Togo), 79 P. ananatis strains from nine countries (Benin, Burkina Faso, Burundi, Mali, Niger, Nigeria, Senegal, Tanzania, Togo), 269 P. stewartii strains from nine countries (Benin, Burkina Faso, Ivory Coast, Mali, Niger, Nigeria, Senegal, Tanzania, Togo) and 220 Pantoea sp. strains from ten countries (Benin, Burundi, Ghana, Ivory Coast, Mali, Niger, Nigeria, Senegal, Tanzania, Togo) (Additional file 1). This result provided first insights on the presence and prevalence of three important Pantoea species in these eleven African countries.
Discussion
Bacterial infections by Pantoea spp. are recognized as being responsible for several diseases of plants, including important crop plants such as rice, maize, sorghum, onion and melon [34–43]. BB of rice caused by species of Pantoea were reported in several countries and include Benin, Togo, Korea, India, Australia, China, Italy, Venezuela, and Russia [10,11,40,44–49].
Given the fact that more than 25 species of Pantoea are currently known and among them several species can infect plants, efficient diagnostic tools are highly demanded by plant pathologists and extension workers. Some plant diseases were attributed to only three species of Pantoea, namely P. agglomerans, P. ananatis and P. stewartii, which can therefore be considered as the major Pantoea species infecting plants. For their diagnosis, several PCR methods are available and have been used but some of them produced amplicons with others species as well [14,50,51], while others are not well reproducible or are inaccessible in typical sub-Saharan laboratory due to specific equipment requirements and/or high costs of some reagents [14,17,18]. Notably, most assays target only one Pantoea species or subspecies. For instance, being of major concern, P. stewartii subsp. stewartii causing Stewart’s bacterial wilt can be detected by several methods but none of them can at the same time identify other bacteria of the genus Pantoea [14,16,18,52,53]. To the best of our knowledge, no robust diagnostic scheme exists that can specifically detect all three major Pantoea species that infect plants.
Based on whole genome sequences, we developed a new multiplex PCR scheme that can specifically detect the three major species of plant-pathogenic Pantoea, P. agglomerans, P. ananatis and P. stewartii. Different strategies can be followed when developing such a multiplex scheme using available whole-genome sequences. One possibility is to automatize the procedure by identifying genomic regions that are shared among a set of strains (e.g. the target species) and which are absent in another set of strains (non-target species). For instance, such an approach was used for the development of a Xanthomonas oryzae-specific multiplex PCR scheme that can differentiate the two pathovars oryzae and oryzicola [9]. The problem with this approach is that it might identify non-essential, often hypothetical genes as targets for the primer design. While present in the training set, it is hard to predict if these non-essential genes will be present and conserved in other, hitherto uncharacterized strains, especially when they originate from other geographical zones and/or belong to more distant genetic lineages.
Here, we targeted housekeeping genes, which are conserved throughout the genus, and relied on lineage (species)-specific sequence polymorphisms. This approach is considered as very robust but it cannot be ruled out that recombination events among strains from different species could undermine the universality of these primer pairs. Yet, we did not find any evidence for such events in any of the sequenced Pantoea strains that were analysed, including environmental isolates and strains isolated from human and plant samples. Nevertheless, because this study was focused on isolates from African rice leaves and seeds and only included a few reference strains from other continents (Additional file 1), it might be of interest to evaluate the new multiplex PCR tool on Pantoea strains isolated from other organisms (others plants, insects, other animals, humans) and from environmental samples.
To reduce the costs and handling time, we generated a multiplex PCR scheme that can work with both purified genomic DNA or with bacterial lysates. In both cases, sufficient specificity and sensitivity were obtained allowing detection of as low as 0.5 ng of DNA or 104 CFU/mL for all three Pantoea species. Such a simple scheme will be of specific interest to phytopathologists, especially in Africa and other less-developed regions. Indeed, diseases due to infections by Pantoea appear to emerge in Africa as recently documented for Benin and Togo [10,11]. In this study, the presence of the three major plant-pathogenic Pantoea species has been demonstrated for eleven African countries. The fact that most of the BB-like symptomatic rice samples proved to contain a high number of Pantoea bacteria suggests that infection by Pantoea is an underestimated source for BB symptoms and might be widespread in Africa. However, more rigorous sampling schemes are required to determine the incidence and prevalence of Pantoea in various rice-growing areas in Africa.
Among the 609 Pantoea isolates, we detected 220 strains (36.1%; additional file 1) of Pantoea sp. that could not be assigned to any of the three species that are specifically targeted by the multiplex PCR scheme. This is an interesting observation that shows that the genus-specific primer pair does not only serve as an internal positive control of the multiplex scheme but that it has its own diagnostic value. Obviously, other species of Pantoea are present in Africa and are likely to cause disease of rice plants as well. Yet, it is still unknown whether or not this group of isolates contains other rice pathogenic species. Pathogenicity assays need to confirm or disprove their status as novel pathogens. Future work will address these isolates, using MLSA and whole genome sequencing.
While screening a large collection of bacterial isolates from rice samples, we also found strains that neither belonged to Pantoea nor to Xanthomonas (data not shown). Some of them were Sphingomonas strains [12], while others may represent new species and genera, which have so far not been connected to rice diseases. These isolates will be further studied by 16S rRNA analysis. From this study, it was concluded that the number of bacterial species that affect rice plants in Africa is certainly larger than previously thought.
Conclusion
A new multiplex PCR scheme was developed to diagnose plant-pathogenic Pantoea spp. This tool enabled the efficient confirmation of the presence of Pantoea species (P. ananatis and P. stewartii) in Benin and Togo, as reported previously, and in several other African countries (Burkina Faso, Burundi, Ghana, Ivory Coast, Mali, Niger, Nigeria, Senegal, Tanzania). Moreover, we found evidence for the presence of P. agglomerans and other species of Pantoea on rice samples from several African countries. This new diagnostic tool will be very useful for crop protection services.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Availability of data and material
Not applicable.
Competing interests
The authors declare that they have no competing interests.
Funding
This publication has been produced with the financial support of the Allocation de Recherche pour une Thèse au Sud (ARTS) program of the Institut de Recherche pour le Développement (IRD). KK received an individual research grant N° C/5921-1 from the International Foundation for Science (IFS). The Africa Rice Center (AfricaRice) and the IRD received financial support from the Global Rice Science Partnership (GRiSP). In addition, AfricaRice received financial support from the Ministry of Foreign Affairs, Japan.
Authors’ contributions
KK and RK conceived and designed the experiments. KK, SD, RA, RD evaluated the primers and multiplex PCR scheme by screening African strains. KK, RK and DS wrote the manuscript. All authors read and approved the final manuscript.
Authors’ information
KK, Institut de Recherche pour le Développement, Montpellier, France; Université de Montpellier, France & Africa Rice Center, Plant Pathology, Cotonou, Benin;
RA, Africa Rice Center, Plant Pathology, Cotonou, Benin;
RD, Africa Rice Center, Plant Pathology, Cotonou, Benin;
DS, Africa Rice Center, Plant Pathology, Cotonou, Benin;
RK, Institut de Recherche pour le Développement, Montpellier, France.
Additional files
Additional file 1: List of bacterial strains used to evaluate the multiplex PCR scheme.
Acknowledgements
We thank Toyin Afolabi (Africa Rice Center Cotonou, Benin), Sandrine Fabre and Florence Auguy (IRD, Cirad, University Montpellier, IPME, Montpellier, France) for excellent technical support. We are grateful to Charlotte Tollenaere for testing this new diagnostic tool in the international laboratory “LMI Patho-Bios” (IRD-INERA Observatoire des Agents Phytopathogènes en Afrique de l’Ouest) in Burkina Faso and for helpful comments on the manuscript.