Detection of Acidovorax avenae subsp . avenae in commercial corn seeds and its correlation with seedling transmission

Acidovorax avenae subsp. avenae (Aaa) caused bacterial leaf streak of corn. To minimize the risk of introducing this pathogen, sorbitol neutral red (SNR) agar was integrated with bio-polymerase chain reaction (PCR) to detect Aaa in commercial corn seeds. A set of 32-random samples was initially detected by sorbitol neutral red (SNR). Colony forming units (CFU) of Aaa from SNR plates and seed extracts was confirmed diagnostic tools by bio-PCR. The bio-PCR produced the positive band from 6, 8 and 10 samples of waxy corn, sweet corn, and field corn with 4, 4, 4, 6, 30, and 40%; 12.5, 14, 16, 23, 30, 30, 30, and 68.2% and 2, 4, 6, 10, 14.3, 20, 20, 30, 30, and 30% respectively, that correlated with the incidence of Aaa detected on SNR agar. We also demonstrated Aaa transmission from sweet corn seeds to seedlings by plant bioassay. As expected, seeds did not show evident symptoms of bacterial leaf streak, but typical Aaa fluid colonies were reisolated indicating seedling transmission of Aaa. The two methods studied provide an effective quantification of Aaa in corns seeds. This suggests that certification of corn seeds against Aaa should be developed.


INTRODUCTION
Acidovorax avenae subsp.avenae (Aaa) is an important pathogen of several hosts including oat, corn, wheat, sugarcane, millet, foxtail, and rice (Kadota, 1996;Prathuangwong et al., 2004).Bacterial leaf streak of corn caused by Aaa was first reported in the major corn growing areas of Thailand by Prathuangwong et al. (2004).Economic losses of 30% caused by Aaa on sweet corn susceptible cultivar, Insee2, have been reported (Techati, 2008).The symptoms of the disease are long streak lesions, water soaking and haloes parallel with leaf vein.These symptoms are similar to other corn diseases such as Stewart wilt caused by Pantoea stewartii subsp.stewartii.
Since the pathogen does not always produce distinct symptoms under field conditions, the disease is very difficult to diagnose (Prathuangwong et al., 2004).The causal agent was reported as a serious seedborne pathogen of various plants such as rice (Shakya, 1987), watermelons, and cantaloupes (Schaad et al., 1978).It can be recovered from seeds using general agar medium (Song et al., 2000;2001) but seedborne contamination of corn was not detected.Moreover, the infected seeds are an important source of primary inoculum.However, Aaa is predominated by other saprophytic bacteria, thus, it is difficult to isolate and identify.Some methods such as blotter test (Shakya and Chung, 1983), serological test (Kadota et al., 1991;Shakya, 1987), and selective media (Kadota, 1996;Zeighler and Alvarez, 1989) have been developed for the detection of Aaa, but none have become widely used.A selective medium would be useful for quantitative recovery and monitoring the pathogen in seed lots.Some selective media (Kadota, 1996;Summer and Schaad, 1977;Zeigler and Alvare, 1989) have limited success in detecting the pathogen directly from seeds or differentiating the bacteria from other seedborne bacteria.Pathogenicity tests are very time-consuming and need further testing to confirm the identity of pathogen.
Plant pathogenic bacteria have been identified quickly by polymerase chain reaction (PCR)-based detection methods (Henson and French, 1993;Song et al., 2004).Although PCR could be a very quick, and clean method for plant pathogenic bacteria detection, it has limitations with non-purified samples such infected seeds or plant leaf samples.Non-purified samples consisted of numerous compounds and saprophytic cells which disturbed the PCR reaction.All these problems were reduced by purification of the target bacteria and extraction of DNA.Moreover, the interfering of classical PCR reaction cause by dead cells and/or free DNA resulted in false positives.
One way to avoid the causal agent of false positives of PCR is to increase the target bacterium on semiselective agar or in liquid media prior to PCR, a technique termed bio-PCR (Schaad et al., 1995;Song et al., 2004).Therefore, the purpose of this study was to combine the use of a modified semi-selective medium, sorbitol neutral red (SNR) with bio-PCR for the specific detection of Aaa in commercial corn seeds.

Bacterial strains and pathogenicity test
Aaa strains were isolated from thirty-four leaf samples of corn seedling at 10 days after planting by routine practice.Virulence of all Aaa strains was assessed on 10 days old corn of susceptible cultivar Insee2, using previously described quantitative methods (Techati, 2008).Briefly, aqueous cell suspensions (OD 600 = 0.2, ca. 10 8 cfu/ ml of each) were sprayed on leaves of plants, covered with plastic bag for 48 h, and maintained in greenhouse.Control plants were inoculated with sterile water in the same manner.The experiment was done in randomized compete block design (RCBD).At four to seven days after inoculation, disease severity was assessed using a scoring method adapted from Pataky et al. (1997) based on the percentage of tissue area affected by Aaa at 3 levels.Disease severity score comprised: A, Short streak lesions and irregular margin, percent of tissue infected area less than 10%; B, long streak lesions, water soaking and haloes parallel with leaf vein, percent of tissue infected area 11-25%; and C, long streak lesions and leaf blight observed, percent of tissue infected area more than 25%.Nine corn leaves, collected from three different plants, were evaluated for each strain.

Recovery of Aaa from infected seeds
The experiment compared the efficiency of the modified SNR [3.0 g of K 2 HPO 4 , 1.0 g of NaH 2 PO 4 , 1.0 g of KNO 3 , 0.2 g of MgSO 4 7H 2 O, 10 ml of neutral red (0.2% aqueous solution 69% active), and 15 g of agar, after autoclaving, 2.0 ml of sterile solution of cyclohexamide (100 mg/ml 75% ethanol), 50 ml of sterile solution of D-sorbitol (10% aqueous)were added, and supplemented with ampicillin 100 µl/mg] in the recovery of Aaa from corn seed extracts with Luria Bertani (LB) agar.20 g of corn seeds from infected and healthy plants, respectively, were surface disinfected before treatment with 95% ethanol for 3 min and washed with sterile distilled water 5 times to remove the 95% ethanol.They were then added to 50 ml of the SNR broth and shaken at 28°C for 15 h.The 10 -1 -10 -3 dilution of infected and non-infected seed extracts and the virulent Aaa strain (Techati, 2008) in LB broth were prepared.0.1 ml of each dilution was inoculated onto ten plates of SNR and LB media by using agar plating method.The plates were incubated at 28°C for 2-4 days and then recovery colony numbers were determined.The experiment was done in RCBD and repeated three times.

Sensitivity of bio-PCR with Aaa
To verify the minimum number of Aaa colonies that could make the PCR positive, representative Aaa strain was inoculated in LB agar.
After 15 h at 28°C, an Aaa colony was removed from the plates using a sterile toothpick and transferred to a PCR tube and 100 µl of sterile water was added.1, 2, 3, and 10 µl of bacterial suspension were spread on LB agar for counting the bacterial concentration.The experiment was done with three replications and was repeated three times.Also, 1, 2, 3, and 10 µl of bacterial suspension in sterile water without DNA extraction were added to the PCR cycles.PCR-amplification was performed with primers AcAVF 5'-GGC TGG ATC ACC TCC TTT C-3' and AcAVR 5'-ACT TGC GAG GTC TTT CAC C-3' designed from 16S-23S internal transcribed spacer (ITS) of Aaa accession number EU368726.1.DNA fragments were amplified in a total volume of 50 μl.The reaction mixture contained 5 μl of 10X ThermoPol rection buffer (100 mM KCl, 10 mM Tris-HCl pH 7.4, 0.1 mM EDTA, 1 mM DTT, 0.5% Tween 20, 0.5% NP-40, and 50% glycerol), 1 μl of 10 pmol of each primer, 2 μl of dNTP mixed, 0.5 μl of Taq DNA Polymerase (BioLabs, USA).A total of 30 amplification cycles were performed in an automated thermocycler (P-100).Each cycle consisted of 15 s of denaturation at 94°C, 30 s of annealing at 58°C, 1 min of extension at 72°C and the last extension step at 72°C was extended to 5 min.Amplified DNA was determined by agarose gel electrophoresis on a horizontal, 0.8% agarose gel in Tris/Borate/EDTA (TBE) buffer (89 mM Tris, 89 mM boric acid, and 2 mM EDTA) at 70 volts constant for 45 min using 1X TBE as a running buffer.The gel was stained with 0.5 μg of ethidium bromide per ml for 10 min and briefly then washed in water before visualizing under UV light and being photographed over a transilluminator (GDS 800, Complete Gel Documentation Analysis System).Moreover, sensitivity of bio-PCR was tested with corn seeds from infected mature plants compared with seeds from healthy plants.20 g seeds were surface disinfected as described above, and added to 50 ml of SNR broth and shaken at 28°C for 15 h.One µl of seed extracts in SNR broth was added to the PCR cycles.PCR condition and reaction was as described above.

Detection of Aaa from naturally contaminated commercial corn seeds
Thirty-two seed samples were randomly collected from commercial corn seeds in Thailand including 6, 9, and 17 seed samples of waxy corn, sweet corn, and field corn, respectively.Fifty seeds from each sample were planted and the disease severity on the leaves at 10 days after planting was observed using the modification method as described above (Pataky et al., 1997).20 g of each seed sample PCR condition and reaction was as described above.The presence of the PCR product was correlated with disease incidence and severity determinations.

DNA sequencing and evaluation
The PCR product was chosen for direct sequencing of the region of 16S-23S ITS.DNA sequencing was done with an Automated DNA Sequencer at the Bioservice Unit, National Science and Technology Development Agency (NSTDA), Thailand.Analysis of sequences was conducted Basic Alignment Search Tool (BLAST) and the DNASTAR Lasergene software package (DNASTAR, Wl, USA).The MegAlign program of the DNASTAR package was used for sequence alignment using CLUSTALW.MegAlignment was then calculated for 16S-23S ITS sequences as available from GenBank for selected Aaa.

Bacterial strains and pathogenicity test
Forty-nine Aaa strains were isolated from leaves of corn seedling (Figure 1).Creamy white, circular smooth with entire margins, glistening Aaa colonies 1-2 mm in diameter on LB medium were isolated consistently from those diseased plants (Figure 2B).Among isolates, Aaa strains were classified by disease severity based on their aggressiveness on Insee2 cultivar into 3 groups.Group A, including 10 weak strains, caused short streak lesions and irregular margins.Group B, including 29 moderately virulent strains, caused long streak lesions, water soaking and haloes parallel with leaf vein.Group C, comprising highly virulent strains, caused long streak lesions and leaf blight.The population of Aaa in this group was 10 strains (Figure 1).

Recovery of Aaa from infected seeds
Modified SNR was the differential medium.Aaa colonies were white, shiny, round, smooth, convex, and 1.0 -1.5 mm in diameter after 4 days and showed different morphology from saprophytic bacteria.Colonies of Aaa on LB were typically 1.0 -2.0 mm diameter after 2 days (Figure 2).The recovery of Aaa on the SNR medium was compared with those on LB.The SNR medium supported good growth of Aaa.The recovery numbers of Aaa ranged from 4x10 3 and 6x10 4 cfu/ml at dilution of 10 -2 and 10 -3 , respectively, on SNR.While no Aaa was detected on LB, saprophytic bacteria of corn seed extracts predominate growth on LB.Also, Aaa was not detected from non-infected seeds by SNR.This indicated that SNR is specific or sensitive enough for reliable detection of Aaa from corn seeds.

Sensitivity of bio-PCR with Aaa
One aim of the research was to verify the efficacy of bio-PCR for Aaa detection with specific primers AcAVF and AcAVR at the various volume of bacterial suspension template.The results showed 1, 2, and 3 µl of Aaa suspension (consisting of 20, 30, and 100 cfu/ml, respectively) made the PCR positive (Figure 3), whereas in the 10 µl sample of Aaa suspension (consisting of 4 x 10 2 cfu/ml), PCR was negative.Moreover, bio-PCR with 1 µl of Aaa suspension template consistently detected Aaa in seeds from infected plants across all samples (100%) and did not detect Aaa in seeds from non-infected plants (100%).This indicated that bio-PCR was able to detect Aaa in corn seeds at concentrations of at least 20 cfu/ml.The sensitivity of bio-PCR with low number of bacterial cells such as 5 to 10 cfu/ml would be developed.

DNA sequencing and evaluation
Alignment analysis of 16S-23S ITS sequence of Aaa isolated from infected corn seeds by MegAlign program (DNAstar inc) revealed that the nucleotide shares identity with 16S-23S ITS sequence (100% at nucleotide levels) of Aaa strain FC-320 (EU368726.1).This confirms that the pathogen isolated from infected corn seeds is Aaa.

DISCUSSION
Among Aaa strains isolated in this study was a heterogeneous species causing varying disease severity on susceptible corn cultivar Insee2.The research supports results reported by Techati (2008) that classified disease severity on corn plants into 3 groups based on the modified disease assessment from Pataky et al. (1997) and correlated with Insee2 inoculated with Aaa.In addition, Insee2 cultivar was the suitable cultivar for classifying Aaa into different groups which correlated with genetic classification by Techati (2008).However, seeds contaminated by Aaa were indistinguishable from healthy seeds based on morphology.Therefore, the best practice for detection of Aaa in corn seeds should be developed.Aaa in infected seeds was detected by SNR with added 100 mg/ml of ampicillin which inhibited growth of other seedborne bacteria where Aaa is resistance to ampicillin.SNR worked well with Aaa corn strains but not with rice and other strains (Song et al., 2000).The bacterial populations recovered from infected seed samples enriched in SNR liquid media and detected by semiselective SNR agar ranged from 4 x10 3 to 6 x 10 4 cfu/ml but were not detected in LB agar.Because of the rapid growth of saprophytic bacteria and other seedborne in LB liquid medium after enrichment for 15 h, colonies of Aaa were not visible when the resulting growth was plated onto LB agar.In contrast, growth of saprophytic bacteria and other seedborne was reduced enough in semiselective SNR agar to allow visual detection of Aaa colonies in most samples when plated onto SNR agars.We confirmed the value of using dsorbitol as the sole carbon source in SNR medium for isolating Aaa of corn in Thailand (Summer and Schaad, 1977).
This SNR media result was correlated with the disease incidence and bio-PCR analyzes.Therefore, both techniques were effective techniques to detect Aaa in corn seeds and could be used for detection of Aaa in commercial seeds.
Corn seeds contaminated with Aaa are important sources of primary inoculum and a means of dissemination of the pathogen to new areas.Although methods including serology (Kadota, 1996;Shakya and Chung, 1983), blotter test and inoculation technique (Techati, 2008) are available for Aaa detection, none are specific or sensitive enough for reliable detection and are time-consuming.Besides being relatively insensitive, serological techniques may detect dead cells or closely related species.In contrast, bio-PCR detected live cells only.The disadvantages of pathogenicity methods or blotter test include time and increased possibilities for cross-contamination.
Based on availability of specific primers, PCR has become a popular technique for identification of bacteria (Schaad et al., 2001).The ITS regions are non-functional elements located between the 16S and 23S genes in prokaryotic rDNA loci and exhibited sequence variation useful for designing specific primers for identifying bacteria at the species level (Neefs et al., 1990;Kostman et al., 1992;Song et al., 1997).Also, the information from this study indicated that AcAVF and AcAVR were specific for Aaa strain from corn.Sensitivity of this bio-PCR technique is high for detecting Aaa with 20 cfu/ml in naturally infected corn seeds.For bio-PCR to be successful, a medium is needed that allows sufficient growth of the target bacterium to allow detection by PCR (generally 1x10 3 cfu/ml or greater) before being overgrown by other bacteria (Schaad et al., 1995).The low number of Aaa that bio-PCR can detect is important for future research.
Therefore, SNR medium and bio-PCR techniques greatly improved the detection of Aaa in commercial corn seeds.Bio-PCR, using SNR medium to enrich the target bacteria in liquid media, allows for rapid and sensitivity for detecting low numbers of the Aaa pathogen in contaminated corn seeds.If high numbers of saprophyte bacteria are encountered after enrichment in solid or liquid media, a DNA extraction step could be included to reduce possible inhibition of PCR.In these cases, the presence of large numbers of other microorganisms prevents easy isolation of the pathogen on semiselective agar.The enrichment time necessary to detect 20 cfu/ml by bio-PCR is only 18 h (15 h enrichment and 3 h PCR reaction).These results show that bio-PCR can be applied to seeds with moderately high levels of background microflora and still detect low levels of Aaa.Furthermore, saprophytic bacteria, overgrow the target bacterium in corn seed samples, and can inhibit the PCR reaction from plant extracts.Bio-PCR using a semiselective medium overcomes these problems and provides a highly sensitive assay for detecting Aaa in comercial corn seeds.The seed assay can be completed within 18 h, which is shorter than the 14-21 days required for traditional culturing, pathogenicity test, and subsequent bacterial identification.
The potential risk of the dissemination of Aaa in corn seeds in the international exchange of corn germplasm is of serious concern.Our results indicates Aaa is seedborne pathogen and transmitted by seed into seedling.SNR medium and bio-PCR protocol as tools greatly improved the detection of Aaa in commercial corn seeds.These results show that SNR and bio-PCR can be applied to seeds with moderately high levels of background microflora and still detect low levels of Aaa.Bio-PCR may be a valuable tool for detecting Aaa in seed lots, monitoring natural bacterial spread, tracking the pathogen in field studies and detecting waterborne or airborne cells caught on filters.

2Figure 1 .Figure 1 .Figure 2 .
Figure 1.Representative of disease severity score (%) of bacterial leaf streak 3 caused by Acidovorax avenae subsp.avenae including, Group A = short 4 streak lesions and irregular margin (A), Group B = long streak lesions, water 5 soaking and haloes parallel with leaf vein 6 A B C Figure 1.Representative of disease severity score (%) of bacterial leaf streak caused by Acidovorax avenae subsp.avenae including, Group A = short streak lesions and irregular margin (A), Group B = long streak lesions, water soaking and haloes parallel with leaf vein (B), Group C = long streak lesions and leaf blight observed (C).

Figure 3 .Figure 4 .
Figure 3. Bio-PCR amplification of the 615-bp fragment of the internal transcribed spacer region of the 16S-23S rDNA of Acidovorax avenae subsp.avenae in infected seed samples with primer AcAVF and AcAVR.PCR products were separated by electrophoresis on a 0.8% agarose gel.Lane 1, 1 kb DNA ladder; lanes 2-7, PCR products amplified; lane 8, 615-bp fragment from genomic DNA of causal agent of bacterial leaf streak.