ionization time-of-flight mass spectrometry

    Direct detection of the plant pathogens Burkholderia glumae, Burkholderia gladioli pv. gladioli, and Erwinia chrysanthemi pv. zeae in infected rice seedlings using matrix assisted laser desorption/ionization time-of-flight mass spectrometry Hideyuki Kajiwara PII: DOI: Reference:

S0167-7012(15)30047-6 doi: 10.1016/j.mimet.2015.08.014 MIMET 4719

To appear in:

Journal of Microbiological Methods

Received date: Revised date: Accepted date:

28 April 2015 19 August 2015 19 August 2015

Please cite this article as: Kajiwara, Hideyuki, Direct detection of the plant pathogens Burkholderia glumae, Burkholderia gladioli pv. gladioli, and Erwinia chrysanthemi pv. zeae in infected rice seedlings using matrix assisted laser desorption/ionization time-of-flight mass spectrometry, Journal of Microbiological Methods (2015), doi: 10.1016/j.mimet.2015.08.014

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Direct detection of the plant pathogens Burkholderia glumae, Burkholderia gladioli

T

pv. gladioli, and Erwinia chrysanthemi pv. zeae in infected rice seedlings using

SC R

IP

matrix assisted laser desorption/ionization time-of-flight mass spectrometry

Hideyuki Kajiwara

National Institute of Agrobiological Sciences, Kannondai 2-1-2, Tsukuba, Ibaraki

MA

NU

305-8602, Japan

Correspondence: National Institute of Agrobiological Sciences, Tsukuba, Ibaraki

D

305-8602, Japan

TE

Tel: +81-29-838-7900; Fax: +81-29-838-7408.

CE P

E-mail: [email protected] (H. Kajiwara)

Keywords: Burkholderia gladioli pv. gladioli, Burkholderia glumae, Erwinia

AC

chrysanthemi pv. zeae, MALDI-TOF MS, Rice (Oryza sativa L.)

1

ACCEPTED MANUSCRIPT

ABSTRACT

T

The plant pathogens Burkholderia glumae, Burkholderia gladioli pv. gladioli, and

by

matrix

assisted

laser

desorption/ionization

time-of-flight

mass

SC R

seedlings

IP

Erwinia chrysanthemi pv. zeae were directly detected in extracts from infected rice

spectrometry (MALDI-TOF MS). This method did not require culturing of the pathogens on artificial medium. In the MALDI-TOF MS analysis, peaks originating

NU

from bacteria were found in extracts from infected rice seedlings. The spectral peaks

MA

showed significantly high scores, in spite of minor differences in spectra. The spectral peaks originating from host plant tissues did not affect this direct MALDI-TOF MS

TE

D

analysis for the rapid identification of plant pathogens.

Highlights

Several pathogens were directly detected in tissue extracts from infected plants.



The detection method was MALDI-TOF MS.



This method is rapid (<5 min) because there is no need to culture the pathogen.

AC

CE P



1. Introduction

Plant pathogens cause many different plant diseases and reduce agricultural crop production, with massive economic losses worldwide every year. Although it is sometimes possible to identify the pathogen based on plant symptoms, it can be difficult to positively identify plant diseases and plant pathogens, especially in rare cases and at the early stages of infection. In the database of plant diseases in Japan, there are 148 kinds of rice diseases, 2

ACCEPTED MANUSCRIPT

including rare cases and physiological conditions (Takeya et al., 2011). Some of them

T

can be discriminated by immunological methods using pathogen-specific antibodies

IP

(Charermroji et al., 2014) or by polymerase chain reaction (PCR) using specific primer

SC R

sets (Vincelli and Tisserat, 2008; Ghyselinck et al., 2011). Mass spectrometric cleaved amplified polymorphic sequence (MS-CAPS) analysis, which can provide results within 1 hour, can also be used for the rapid discrimination of plant pathogens (Kajiwara et al,

NU

2012; Kajiwara 2015). However, before analyzing the causal pathogen, suitable

MA

antibodies or primer sets must be chosen based on the plant symptoms. The pre-selection of the candidate(s) among the 148 kinds of plant diseases in rice relies on

D

the experience of the analysts.

TE

Bacterial panicle blight of rice is caused by Burkholderia glumae (Azegami, 2009; Riera-Ruiz et al., 2014) and Burkholderia gladioli pv. gladioli (Cother et al.,

CE P

2010). This plant disease occurs worldwide (Ham et al., 2011). Both pathogens cause almost the same symptoms on panicles, so it is difficult to discriminate between them. A

AC

previous study reported the use of PCR to discriminate between these pathogens (Maeda et al., 2006). The two pathogens show more than 99% similarities in their 16S ribosome RNA internal transcribed spacer sequences (Nandakumar et al., 2009). Erwinia chrysanthemi pv. zeae is another pathogen of rice that causes bacterial foot rot and seriously damages rice crops (Wensing et al., 2011). Matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) techniques were first applied in the detection of microorganisms for clinical purposes (Buchan and Ledeboer, 2014) and were expected to be useful for analyzing plant pathogens (Ahmad et al., 2012). According to recent reports, the MALDI-TOF MS technique, which has been named MALDI biotyping, is more than 3

ACCEPTED MANUSCRIPT

95% accurate in medical microorganism discrimination (Spanu et al., 2012). However,

T

there are two main problems in using MALDI-TOF MS to detect plant diseases. First,

IP

there is insufficient mass fingerprint data of plant pathogens in databases. To address

SC R

this problem, I have started a collection of MS data for plant pathogens using information in the genebank of the National Institute of Agrobiological Sciences (NIAS). The other main problem is that pathogens need to be cultured on a suitable medium for

NU

1–2 days to obtain colonies for MALDI-TOF MS analysis.

MA

This paper describes direct MALDI biotyping using extracts from infected rice seedlings for the rapid discrimination of plant pathogens. B. glumae, B. gladioli, and E.

D

chrysanthemi could be directly identified from extracts of infected rice seedlings by the

TE

mass fingerprints obtained in MALDI-TOF MS analyses. This is the first report of the

CE P

direct detection of plant pathogens in plant tissue extracts by MALDI-TOF MS analysis.

AC

2. Materials and methods

2.1. Bacteria

B. glumae (MAFF 301682), B. gladioli pv. gladioli (MAFF 302386), and E. chrysanthemi pv. zeae (MAFF 106501) were obtained from the genebank at the NIAS. Each strain was cultured on Yeast extract- Malt extract (YM) agar plates (4 g/l yeast extract (Nakalai Tesque, Kyoto, Japan), 10 g/l malt extract (Nakalai Tesque), 4 g/l glucose (Wako Pure Chemical Industries, Osaka, Japan), 18 g/l agar (Nakalai Tesque). pH 7.3) at 25 C or in YM liquid medium for 2 days. Bacteria in YM liquid medium were harvested by centrifugation at 15,000 rpm for 5 min and washed three times with the same volume of distilled sterilized water (DSW). Bacterial suspension (50 l) was 4

ACCEPTED MANUSCRIPT

T

added to 1-week-old seedlings.

IP

2.2. Plant materials

SC R

Rice (Oryza sativa L. cv. Nipponbare) seeds were sterilized in 70% ethanol for 30 s and then immersed in 50% (v/v) Kitchen Haiter (Kao Corporation, Tokyo, Japan) containing 6% (w/v) sodium hypochlorite for 10 min, with gentle stirring. After washing three

NU

times with DSW, single seeds were placed in test tubes containing 3 ml Murashige and

MA

Skoog plant salt medium (Wako) without sugar or vitamins. Plants were grown at 25 C

D

under constant illumination.

TE

2.3. Sample preparation

Each infected or healthy rice seedling (3 weeks old) was washed three times with DSW

CE P

and then ground in a tube with 500 l DSW. Extracts were passed through a 1,000-l tip, which contained a 1 × 1 cm piece of Miracloth (Calbiochem, La Jolla CA, USA). The mixture was centrifuged at 15,000 rpm for 5 min, and the pellet was suspended in 300

AC

l DSW by vortexing. Then, 900 l ethanol (Nakalai Tesque) was added, and the mixture was centrifuged as described above. The pellet was used for MALDI-TOF MS analysis. Samples were stored at −20 C until use.

2. 4. MALDI-TOF MS analysis Bacteria were treated with 70% (v/v) formic acid (Wako) and the same volume of acetonitrile (Wako), and then centrifuged at 15,000 rpm for 5 min. The pellet was used for analysis. Extracts (0.5 l) were spotted on the 384 stainless MALDI-TOF MS target (Bruker Daltonics, Billerica MA, USA) and dried at room temperature. Then, the same 5

ACCEPTED MANUSCRIPT

spot was overlaid with 0.5 l -cyano-4-hydroxycinnamic acid (Fluka, St. Louis MO,

T

USA) saturated in 50:47.5:2.5 acetonitrile:water:trifluoroacetic acid (Nakalai Tesque).

IP

The Ultraflex instrument (Bruker Daltonics) was used in linear, positive mode. Ions

SC R

were generated by a 20-Hz nitrogen laser. Each spectrum resulted from 50 laser shots at 20 random positions within the spots. Mass spectra were generated in the mass range of m/z 2,000 to 20,000. Calibration was performed with an MBT standard (Bruker

NU

Daltonics). Mass spectra were analyzed by Flex Analysis 2.0 and MALDI Biotyper OC

MA

3.1 using the default parameters in the Biotyper processing standard method (Bruker

D

Daltonics).

TE

3. Results

CE P

3.1. Analysis of healthy seedlings

Extracts from the infected rice seedlings contained a large amount of plant tissue debris.

AC

Because these tissues interfered with the MALDI-TOF MS analysis, they had to be removed from the extracts. A piece of Miracloth inserted into the point of a 1,000-l tip effectively removed most of the plant tissue. After the subsequent centrifugation step, the pellet contained small plant tissue debris, insoluble metabolites, and cell organs including chloroplasts, as well as the pathogenic bacteria. Ethanol was added to rupture the chloroplasts. Proteins in rice cells and chloroplasts remained in supernatant after the next centrifugation step. The main component of the precipitate was assumed to be the pathogenic bacterial cells. To determine whether the Miracloth affected the results, spectra were compared between Escherichia coli cells prepared with or without the Miracloth step. There was 6

ACCEPTED MANUSCRIPT

no difference between the spectra; i.e., no significant differences in the mass

T

fingerprints of E. coli prepared with and without Miracloth (data not shown).

IP

A syringe filter (0.22 m) was tested to determine whether it could trap the

SC R

bacteria in the extract on the membrane after passing the extract through Miracloth. A similar step was included in PCR experiments in another study (Ohara et al., 2010). Although the bacteria were trapped on the filter, the pressure became very high because

NU

of small plant tissue debris. Extracts eluted with formic acid and acetonitrile could be

MA

analyzed by MALDI-TOF MS (data not shown). However, the concentration of the pathogen was low in these eluted solutions, and sometimes the syringe filter broke

D

because of the high pressure (data not shown).

TE

Before analyzing tissues from pathogen-infected rice seedlings, extracts from healthy rice seedlings were analyzed by MALDI-TOF MS to determine the background

CE P

spectrum (Fig. 1). The major peak in the spectrum of healthy rice seedlings was at m/z 3,334, and there were no other significant mass fingerprints. Therefore, overlaps

AC

between spectral peaks derived from host plants and those derived from the pathogen were not considered to affect the MALDI biotyping analysis.

3.2. Analysis of infected rice seedlings Extracts from rice seedlings infected with B. glumae (Fig. 2A), B. gladioli (Fig. 3A), and E. chrysanthemi (Fig. 4A) were analyzed by MALDI-TOF MS. The infected rice seedlings had a scorched appearance and brown withered leaves. Most bacteria on the surface of the seedling leaf were removed in the DSW washing step. Therefore, the peaks in the spectra were considered to have originated from bacteria infecting the host plant tissue. The major peak observed in the spectrum of healthy seedlings (m/z 3,334) 7

ACCEPTED MANUSCRIPT

was absent from the spectrum of infected rice seedlings.

T

The mass fingerprints obtained from B. glumae (Fig. 2B), B. gladioli (Fig. 3B),

IP

and E. chrysanthemi (Fig. 4B) grown on YM agar plates showed many peaks. Most of

SC R

them were consistent with the peaks in the spectra of seedlings infected with these pathogens. However, the mass fingerprints showed some differences between pathogen-infected seedlings and the corresponding pathogens grown on YM agar plates.

NU

The spectrum of plants infected with B. glumae showed peaks at m/z 4,367, m/z

MA

4,408, and m/z 5,195, but these peaks were absent from the mass fingerprints of the pathogens grown on YM medium. The spectrum of B. glumae grown on agar plates

D

showed peaks at m/z 2,070, m/z 2,501, m/z 4044, m/z 7,097, m/z 8,891, and m/z 9,293.

TE

The spectrum of B. gladioli grown on agar plates showed peaks at m/z 2,438, m/z 5,103, and m/z 6,751, but these were absent from the spectrum of plants infected with this

CE P

pathogen. The spectrum of plants infected with B. gladioli had a peak at m/z 7,805, which was absent from the spectrum of B. gladioli grown on agar plates. The spectrum

AC

of rice plants infected with E. chrysanthemi showed peaks at m/z 6,861, m/z 8,131, and m/z 15,084. The spectrum of E. chrysanthemi grown on YM medium showed peaks at m/z 7,028, m/z 10,936, and m/z 12,280. In many instances, the heights of common peaks differed between the spectrum of the pathogen and the spectrum of the plant infected with the pathogen. Data analysis of B. glumae, B. gladioli, and E. chrysanthemi using the MALDI biotyper gave scores of 2.203, 2.089, and 2.234, respectively. These scores showed more than 95% accuracy of matching, despite the peaks from the host plant in the spectrum.

8

ACCEPTED MANUSCRIPT

T

4. Discussion

IP

The spectrum of healthy seedlings, but not infected seedlings, had a peak at m/z 3,334

SC R

(Fig. 1). This peak was not considered to represent a protein, because no band was observed in a high-concentration sodium dodecylsulfate polyacrylamide gel electrophoresis analysis (data not shown). Further experiments were conducted to

NU

analyze this peak in more detail. The leaves used for the direct analysis were brown.

MA

Thus, it is likely that most of the cells in the infected seedling tissue were dead. This would explain why the substance disappeared from the mass fingerprints of infected

D

seedlings.

TE

The mass fingerprints differed between infected seedlings and pathogens grown on agar media. For example, the peaks at m/z 4,367, m/z 4,408, and m/z 5,195 that were

CE P

present in the spectrum of B. glumae-infected seedlings were absent from the mass fingerprints of B. glumae cultured on YM agar plates. These peaks were considered to

AC

represent products of the host plant or the pathogen resulting from the plant–pathogen interaction, but there is no evidence for what caused those minor differences. It is important to collect mass fingerprint data for plant pathogens in their host plants for more accurate discrimination using the direct MALDI biotyping technique. The MALDI biotyping method has been used for medical purposes. In such analyses, bacterial colonies must be obtained by culture at 37 C for at least 1 day. In the case of plant pathogens, the culture temperature is lower and the growth time longer, usually 2–3 days are required to obtain colonies on agar plates. Most plant pathogens can grow on YM or potato dextrose media, although some require special culture media and conditions. There have been several reports on the MALDI biotyping of plant 9

ACCEPTED MANUSCRIPT

pathogens grown on artificial media. Those pathogens include Erwinia (Sauer et al.,

T

2008; Wensing et al., 2011), Verticillium (Tao et al., 2009), Clavibacter (Zaluga et al.,

IP

2011), Fusarium (Dong et al., 2009; Kemptner et al., 2009), nematodes (Perera et al.,

SC R

2005), and microorganisms in soil and roots (Hahn et al., 2011; Sets et al., 2013). As far as I know, there have been no reports of MALDI biotyping to directly discriminate plant pathogens in plant extracts. Direct MALDI biotyping eliminates the need to culture the

NU

pathogen, which usually takes 2–3 days. At present, however, not all plant pathogens

MA

can be directly detected in plant tissue extracts. For the rapid discrimination of plant diseases and plant pathogens, mass fingerprint spectra of infected plants and pathogens

D

must be collected one by one and added to a database.

TE

There are three main advantages in detecting plant pathogens directly from plant tissue extracts. First, there is no need to culture the pathogen, allowing rapid

CE P

identification. Culturing plant pathogens on agar plates requires a few days. If there is a practical method to detect the plant pathogen using MALDI-TOF MS analysis, it takes

AC

only 5 min to detect the pathogen from approximately 1 cm2 of the infected region. The second advantage is that pre-selection of the plant pathogens prior to analysis is not required. Therefore, the method does not rely on the plant pathology experience of the operator. For PCR or immunochemical analyses of plant pathogens, the operator must first select suitable primer sets or antibodies for candidate pathogens. Analyses using PCR are among the most reliable methods, as long as specific primer sets are available. However, such analyses require 2–3 h for the amplification and 1–2 h for agarose electrophoresis. Previously, I developed an MS-CAPS method for the rapid identification of pathogens. This MS-CAPS method took nearly 1 h from sampling to data analysis (Kajiwara et al, 2012; Kajiwara 2015), but it required the selection of 10

ACCEPTED MANUSCRIPT

suitable primer sets based on the disease symptoms. If the causal agent of the disease is

T

not apparent based on the symptoms in plants, several experiments must be performed

IP

to identify candidate pathogens. Also, because the genomes of most plant pathogens

SC R

have not yet been fully sequenced, it is often difficult to produce pathogen-specific primer sets. For these reasons, analyses based on PCR or immunochemistry are still of limited value in plant pathology.

NU

The third advantage of this new method is the low running cost. Direct analysis of

MA

the pathogen from infected tissue using MALDI-TOF MS does not require bacterial cultures, PCR amplification, or immunochemical analyses. The present method has

D

some manual operations and consumes some chemicals. Although the extraction of

TE

pathogens from infected plant tissues must be performed manually in the first step of the MALDI biotyping method, the other steps in the process, including spotting the

CE P

sample on the target and data analysis, can be performed automatically using machines. In other words, many samples can be analyzed simultaneously.

AC

As far as I know, this is the first report of direct MALDI biotyping analysis of plant pathogens from infected plant materials. B. glumae, B. gladioli, and E. chrysanthemi could be directly detected by MALDI biotyping. Rice seedlings were also grown with other bacteria; Acidovorax avenae subsp. avenae (MAFF 106618, bacterial brown stripe), Burkholderia plantarii (MAFF 301723, bacterial seedling blight), Pseudomonas fuscovaginae (MAFF 106631, sheath brown rot), Pseudomonas fluorescens (MAFF 106531, not proposed), and Pseudomonas syringae pv. aptata (MAFF 302830, bacterial grain rot). These bacteria showed specific mass fingerprints when grown on YM agar plates (data not shown). However, none of them showed identical peaks in the spectra of extracts from infected rice seedlings. This may be 11

ACCEPTED MANUSCRIPT

because the pathogen did not successfully invade the plant tissues, or because there was

T

an insufficient bacterial load in the extracts. Therefore, this direct method to identify

IP

plant pathogens by MALDI biotyping is in the first stage of development. Further

SC R

investigations in collaboration with plant pathologists are required to refine and improve the method.

The direct detection of pathogens using MALDI biotyping has been demonstrated

NU

in ideal conditions. In field environments, resident bacteria are present on the surface or

MA

inside of plant tissues. The area showing symptoms on diseased plants is assumed to be dominated by the pathogen, which would have out-competed resident bacteria, or at

D

least affected the balance between resident and pathogenic bacteria during infection.

TE

Therefore, it is expected that the signals derived from plant pathogens would still be detectable in field samples by MALDI biotyping.

CE P

There are still many problems in the direct detection of plant pathogens from infected plant tissue. Resident bacteria are one of the most serious problems; however, if

AC

the mass spectrum of the resident bacteria is known, it can be subtracted from the spectral data. The mass spectrum of the host plant tissue should be also subtracted from the MALDI biotyping data. Such corrections will increase sensitivity of the direct MALDI biotyping technique. The program MALDI biotyper does not yet have this function. The software Speclust (Aim et al., 2006), which is available online, allows the peak masses of host plants to be deleted manually, although this software does not allow analyses of peak heights. Removing the resident bacteria and substances that cause undesired peaks in the spectra of plant tissues would contribute to reducing noise in the MALDI-TOF MS analysis. Although the peak observed at m/z 3,334 was not identified, antibodies against 12

ACCEPTED MANUSCRIPT

the corresponding substance would be helpful for this purpose. In the same way,

T

antibodies against resident bacteria would be useful for reducing noise.

IP

The mass spectra data for the pathogens detected in this study have been collected

SC R

and added to a database. Although the database for MALDI biotyping has recently been expanded, it contains few microorganisms that cause plant disease, and so it has limited use for clinical purposes. The NIAS has a genebank that contains information for some

NU

plant pathogens (Takeya et al., 2011). Further collection of mass fingerprint data for

MA

plant pathogens and refinements to the direct MALDI biotyping method will contribute

D

to making it a practical analytical technique for agriculture.

TE

5. Conclusion

CE P

The plant pathogens B. glumae, B. gladioli pv. gladioli, and E. chrysanthemi could be detected in extracts from infected rice seedlings using MALDI-TOF MS. Whereas

AC

previously reported MALDI biotyping methods use bacterial colonies to identify plant pathogens, this new method allows the direct identification of pathogens in extracts from infected host plants. This eliminates the need to culture the pathogen, saving both time and resources.

Acknowledgement

I thank Dr. T. Aoki for initial support with bacteria handling.

13

ACCEPTED MANUSCRIPT

T

References

IP

Ahmad, F., Babalola, O.O., Tak, H.I., 2012. Potential of MALDI-TOF mass

SC R

spectrometry as rapid detection techniques in plant pathology: identification of plant-associated microorganisms. Anal. Bioanal. Chem. 404, 1247-1255. Aim, R., Johansson, P., Hjerno, K., Emanuelsson, C., Ringner, M., Hakkinen, J., 2006.

NU

Detection and identification of protein isoforms using cluster analysis of

MA

MALDI-MS mass spectra. J. Proteome Res. 5, 785-792. Azegami, K., 2009. Burkholderia glumae and Burkholderia plantarii, the pathogens of

D

bacterial grain rot of rice and bacterial seedling blight of rice. MAFF

TE

Microorganism Genetic Resources Manual 26. (Japanese) Buchan, B.W., Ledeboer, N.A., 2014. Emerging technologies for the clinical

CE P

microbiology laboratory. Clin. Microbiol. Rev. 27, 783-822. Charermroj, R., Hinananto, O., Seepiban, C., Kumpoosiri, M., Warin, N., Gajanandana,

AC

O., Elliott, C.T., Karoonuthaisiri, N., 2014. Antibody assay in a multiwall plate format for the sensitive and multiplexed detection of important plant pathogen. Anal. Chem. 86, 7049-7056. Dong, H., Kemptner, J., Marchetti-Deschmann, M., Kubicek, C.P., Allmaier, G., 2009. Development of a MALDI two-layer volume sample preparation technique for analysis of colored conidia spores of Fusarium by MALDI linear TOF mass spectrometry. Anal. Bioanal. Chem. 395, 1373-1383. Cother, E.J., Noble, D.H., de Ven, R.J.V., Lanoiselet, V., Ash, G., Vuthy, N., Visarto, P., Stodart, B., 2010. Bacterial pathogens of rice in the kingdom of Cambodia and description of a new pathogen causing a serious sheath rot disease. Plant Pathol. 59, 14

ACCEPTED MANUSCRIPT

944-953.

T

Fory, P.A., Triplett, L., Balle, C., Abello, J.F., Duitama, J., Aricapa, M.G., Prado, G.A.,

of two emerging rice seed bacterial pathogens.

SC R

Comparative analysis

IP

Correa, F., Haamilton, J., Leach, J.E., Tohme, J., Mosquera, G.M., 2014.

Phyotopathology 104, 436-444.

Ghyselinck, J., van Hoorde, K., Hoste, B., Heylen, K., De Vos, P., 2011. Evaluation of

NU

MALDI-TOF MS as a tool for high-throughput dereplication. J. Microbiol.

MA

Methods 86, 327-336.

Hahn, D., Mirza, B., Benagli, C., Vogel, G., Tonolla, M., 2011. Typing of

D

nitrogen-fixing Frankia strains by matrix-assisted laser desorption/ionization

TE

time-of-flight (MALDI-TOF) mass spectrometry. Syst. Appl. Microbiol. 34, 63-68. Ham, J.H., Melanson, R.A., Rush M.C., 2011. Burkholderia glumae; next major

CE P

pathogen of rice? Mol. Plant Pathol. 12, 329-339. Kajiwara, H., 2015. Gene analysis using mass spectrometric cleaved amplified

AC

polymorphic sequence (MS-CAPS) with matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF). Methods Mol. Biol. 1245, 205-214.

Kajiwara, H., Sato, M., Suzuki, A., 2012. Detection of Acidovorax avenae subsp. citrulli using PCR and MALDI-TOF MS. J. Electrophoresis 56, 13-17. Kemptner, J., Marchetti-Deschmann, M., Mach, R., Druzhinina, I.S., Kubicek, C.P., Allmaier, G., 2009. Evaluations of matrix-assisted laser desorption/ionization (MALDI) preparation techniques for surface characterization of intact Fusarium spores by MALDI linear time-of-flight mass spectrometry. Rapid Commun. Mass spectrum. 23, 877-884. 15

ACCEPTED MANUSCRIPT

Maeda, Y., Shinohara, H., Kiba, A., Ohnishi, K., Furuya, N., Kawamura, Y., Ezaki, T.,

T

Vandamme, P., Tsushima, S., Hikichi, Y., 2006. Phylogenic study and multiplex

IP

PCR-based detection of Burkholderia plantarii, Burkholderia glumae and

SC R

Burkholderia gladioli using gyrB and rpoD sequences. Int. J. Syst. Evol. Microbiol. 56, 1031-1038.

Nandakumar, R., Shahjahan, A.K.M., Yuan, X.L., Dickstein, E.R., Groth, D.E., Clark,

NU

C.A., Cartwright, R.D., Rush, M.C., 2009. Burkholderia glumae and Burkholderia

MA

gladioli cause bacterial panicle blight in rice in the Southern United States. Plant Dis. 93, 896-905.

D

Ohara, T., Sawada, H., Azegami, K., 2004. Detection of the pathogen of bacterial grain

TE

rot of rice from agro-environment using membrane filtration and enrichment culture followed by PCR assay. Jpn. J. Phytopathol. 70, 115-122. (Japanese)

CE P

Perera, M.R., Vanstone, V.A., Jones, M.G.K., 2005. A novel approach to identify plant parasitic nematodes using matrix-assisted laser desorption/ionization time-of-flight

AC

mass spectrometry. Rapid Commun. Mass spectrum. 19, 1454-1460. Riera-Ruiz, C., Vargas, J., Cedeno, C., Quirola, P., Escobar, M., Cevallos-Cevallos, JM., Ratti, M., Perath, E.L., 2014. First report of Burkholderia glumae causing bacterial panicle blight on rice in Ecuador. Plant Dis. 98, 988-989. Sauer, S., Freiwald, A., Maier, T., Kube, M., Reinhardt, R., Kostzewa, M., Geider, K., 2008. Classification and identification of bacteria by mass spectrometry and computational analysis. PlosOne 3, e2843. Spanu, T., Posteraro, B., Fiori, B., D’Inzeo, T., Campoli, S., Ruggeri, A., Tumbarello, M., Canu, G., Trecarichi, E.M., Parisi, G., Tronci, M., Sanguinetti, M., Fadda, G., 2012. Direct MALDI-TOF mass spectrometry assay of blood culture broths for 16

ACCEPTED MANUSCRIPT

rapid identification of Cardida species causing bloodstream infections: an

T

observational study in two large microbiology laboratories. J. Clin. Microbiol. 50,

IP

176-179.

SC R

Stets, M.I., Pinto Jr., A.S., Huergo, L.F., de Souza, E.M., Guimaraes, V.F., Alves, A.C., Berenice, M., Steffens, R., Monterio, R.A., Pedrosa, F.O., Cruz, L.M., 2013. Rapid identification of bacterial isolated from wheat roots by high resolution whole cell

NU

MALDI-TOF MS analysis. J. Biotechnol. 165, 167-174.

MA

Takeya, M., Yamasaki, F., Uzuhashi, S., Aoki, T., Sawada, H., Nagai, T., Tomioka, K., Tomooka, N., Sato, T. Kawase, M., 2011. NIASGBdb: NIAS genebank databases

D

for genetic resources and plant disease information. Nucleic Acids Res. 39, (Suppl.

TE

1): D1108-D1113.

Tao, J., Zhang, G., Jiang, Z., Cheng, Y., Feng, J., Chen, Z., 2009. Detection of

CE P

pathogenic Verticillium spp. using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun. Mass spectrum. 23, 3647-3654.

AC

Vincelli, P., Tisserat, N. 2008. Nucleic acid-based pathogen detection in applied plant pathology. Plant Dis. 92, 660-669. Wensing, A., Gernold, M., Geider, K., 2011. Detection of Erwinia species from the apple and pear flora by mass spectrometry of whole cells and with novel PCR primers. J. Appl. Microbiol. 112, 147-158. Zaluga. J., Heylen, K., VanHoorde, K., Hoste, B., van Vaerenbergh, J., Maes, M., de Vos, P., 2011. GyrB sequences analysis and MALDI-TOF MS as identification tools for plant pathogenic Clavibacter. Syst. Appl. Microbiol. 34, 400-407.

17

ACCEPTED MANUSCRIPT

T

Figure legends

IP

Fig. 1. Matrix assisted laser desorption/ionization time-of-flight mass spectrometry

AC

CE P

TE

D

MA

NU

healthy seedlings showed a peak at m/z 3,334.

SC R

analysis of healthy rice seedlings grown on Murashige and Skoog medium. Spectra of

18

ACCEPTED MANUSCRIPT

T

Fig. 2. Matrix assisted laser desorption/ionization (MALDI) time-of-flight mass

IP

spectrometry analysis of B. glumae. MALDI biotyping was performed on extracts from

SC R

rice seedlings infected with B. glumae and bacteria grown on Yeast extract- Malt extract (YM) plates. (A) Mass spectrum of rice seedling infected with B. glumae. Peaks at m/z 4,367, m/z 4,408, and m/z 5,195 were specific to infected seedlings. (B) Mass spectrum

NU

of B. glumae grown on YM plates. Peaks at m/z 2,070, m/z 2,501, m/z 4,044, m/z 7,097,

AC

CE P

TE

D

MA

m/z 8,891, and m/z 9,293 were specific to cultured bacteria.

19

ACCEPTED MANUSCRIPT

T

Fig. 3. Matrix assisted laser desorption/ionization (MALDI) time-of-flight mass

IP

spectrometry analysis of B. gladioli. MALDI biotyping was performed on extracts from

SC R

rice seedlings infected with B. gladioli and bacteria grown on Yeast extract- Malt extract (YM) plates. (A) Mass spectrum of rice seedling infected with B. gladioli. Peak at m/z 7,805 was specific to infected seedlings. (B) Mass spectrum of B. gladioli grown on YM

AC

CE P

TE

D

MA

NU

plates. Peaks at m/z 2,438, m/z 5,103, and m/z 6,751 were specific to cultured bacteria.

20

ACCEPTED MANUSCRIPT

T

Fig. 4. Matrix assisted laser desorption/ionization (MALDI) time-of-flight mass

IP

spectrometry analysis of E. chrysanthemi. MALDI biotyping was performed on extracts

SC R

from rice seedlings infected with E. chrysanthemi and bacteria grown on Yeast extractMalt extract (YM) plates. (A) Mass spectrum of infected rice seedling with E. chrysanthemi. Peaks at m/z 6,861, m/z 8,131, and m/z 15,084 were specific to infected

NU

seedlings. (B) Mass spectrum of E. chrysanthemi grown on YM plates. Peaks at m/z

AC

CE P

TE

D

MA

7,028, m/z 10,936, and m/z 12,280 were specific to cultured bacteria.

21

ACCEPTED MANUSCRIPT

Highlights Several pathogens were directly detected in tissue extracts from infected plants.



The detection method was MALDI-TOF MS.



This method is rapid (<5 min) because there is no need to culture the pathogen.

AC

CE P

TE

D

MA

NU

SC R

IP

T



22