Search PMN  

 

PDF version
for printing

Peer Reviewed
Impact
Statement




2004 Plant Management Network.
Accepted for publication 31 March 2004. Published 5 April 2004.


Discrimination of Plant Pathogenic Bacteria Using an Electronic Nose


M. T. Momol, North Florida Research & Education Center, 155 Research Road, Quincy 32351; M. O. Balaban, F. Korel, and A. Odabasi, Food Science and Human Nutrition Department, P.O. Box 110370, University of Florida, Gainesville 32611-0370; E. A. Momol, North Florida Research & Education Center, 155 Research Road, Quincy 32351; G. Folkes, Food Science and Human Nutrition Department, P.O. Box 110370, University of Florida, Gainesville 32611-0370; and J. B. Jones, Plant Pathology Department, P.O. Box 110680, University of Florida, IFAS, Gainesville 32611


Corresponding author: Timur (Tim) Momol. tmomol@ufl.edu


Momol, M. T., Balaban, M. O., Korel, F., Odabasi, A., Momol, E. A., Folkes, G., and Jones, J. B. 2004. Discrimination of plant pathogenic bacteria using an electronic nose. Online. Plant Health Progress doi:10.1094/PHP-2004-0405-01-HN.


The term electronic nose (EN) is used to describe such an array of chemical sensors, where each sensor has only fractional specificity to a wide range of odor molecules, connected with a suitable pattern recognition system. Sensor technology based on conducting polymers, quartz-resonator sensors, and others are used in food processing industry to improve quality control techniques (1). A rapid, sensitive, specific, nondestructive, and easy-to-use technique such as the EN could be utilized for detection and identification of plant pathogenic bacteria in plant diagnostic clinics and quarantine laboratories. The use of the EN for the identification of plant pathogens was reported previously as abstracts (3,4).

The discrimination of seven species of plant pathogenic bacteria (Acidovorax avenae subsp. citrulli, Agrobacterium tumefaciens, Clavibacter michiganensis subsp. michiganensis, Erwinia amylovora, Pseudomonas syringae pv. tomato, Ralstonia solanacearum, and Xanthomonas campestris pv. vesicatoria) by measuring the volatile compounds produced from pure cultures has been performed using an EN and Discriminant Function Analysis (DFA) (2). The identity of all strains was confirmed by fatty acid methyl esters profile data analyses using the Microbial Identification System (MIDI Inc., Newark, DE). All bacteria were grown on Trypticase soy agar (trypticase soy broth [BBL] containing Bacto Agar [Difco]) poured in 60-mm glass Petri plates at 28C for 24 2 h. Before incubation, plates were sealed with parafilm until processing. Each species or strain was grown in a separate plate. For all experiments, the same number (four or five replications) of bacteria-free control media samples were prepared and subjected to the same conditions as bacterial cultures. An EN (e-Nose 4000, EEV Inc., Elmsford, NY) was used for odor measurements of the bacterial species. Before each experiment, the electronic nose was calibrated with a polypropylene glycol solution (75% v/v water). Four or five samples (replicates) of each bacterial species and the control sample (uninoculated media) were read once separately with the EN. Experiments were conducted by using two species, two strains of two species, five species, and seven species of plant pathogenic bacteria during the same run and bacteria-free control media. For each bacterial species and each combination, the same experiment was repeated twice. Analysis of electronic nose sensor data of bacterial species and the control was performed with Statistica (StatSoft Inc., Tulsa, OK) using DFA (2). The correct classification rate for two bacterial species at a time (i.e., Xanthomonas campestris pv. vesicatoria versus Pseudomonas syringae pv. tomato) and control samples was 100.0% (Fig. 1) for each combination. The correct classification rate for two strains per bacterial species and control samples was 100.0% (Fig. 2). The correct classification rate for five bacterial species and bacteria-free control samples was 97.0% and 96.67% in two different experiments (data not shown). When volatiles from seven bacterial species and the control were compared, the correct classification rate was 90.0 (Fig. 3), 93.3, and 95.0% in three different runs using DFA.


 

Fig. 1 Discriminant function analysis of two bacterial species based on electronic nose readings. Abbreviations used: Control = bacteria free, XV = Xanthomonas campetris pv. vesicatoria, PT = Pseudomonas syringae pv. tomato. (Click image for larger view.)

 

Fig. 2. Discriminant function analysis of two bacterial strain per species, based on electronic nose readings. Abbreviations used: RS1 = Ralstonia solanacearum strain 1, RS2 = Ralstonia solanacearum strain 2, XV91-118 = Xanthomonas campestris pv. vesicatoria strain 91-118, XV92-7 = Xanthomonas campestris pv. vesicatoria strain 92-7. (Click image for larger view.)


 

Fig. 3. Discriminant function analysis of seven bacterial species based on electronic nose readings. Abbreviations used: Control = bacteria free, RS = Ralstonia solanacearum, EA = Erwinia amylovora, AT = Agrobacterium tumefaciens, AAC = Acidovorax avenae subsp. citrulli, CLA = Clavibacter michiganensis subsp. michiganensis, PT = Pseudomonas syringae pv. tomato, XV = Xanthomonas campestris pv. vesicatoria. (Click image for larger view.)

 

In this study, the EN with 12 polymer sensors discriminated between plant pathogenic bacteria belonging to up to seven species from seven different genera. DFA analysis proved to be successful in performing overall correct classification (90.0%) of the seven bacterial species used in these experiments. Correct classification rate was higher if five or two species were tested in the same run. The EN technology is novel and in its infancy for application in plant pathology. This study is reported to encourage plant pathologists in industry and public sectors to explore EN technologies for plant pathology applications.


Literature Cited

1. Bartlett, P. N., Elliot, J. M., and Gardner, J. W. 1997. Electronic noses and their application in the food industry. Food Technol. 51:44-48.

2. Maul, F., Sargent, S. A., Huber, D. J., Balaban, M. O., Sims, C. A., and Baldwin, E. A. 1999. Harvest maturity and storage temperature affect volatile profiles of ripe tomato fruits: Electronic nose and gas chromatographic analyses. Pages 1-13 in: Electronic Noses and Sensor Array Based Systems Design and Applications. W. J. Hurst, ed. Technomic Publishing Co., Inc. Lancaster, PA.

3. Momol, M. T., Halsey, L., Fletcher, J., Balaban, M. O., Zazueta, F., and Kucharek, T. A. 2000. Evaluation of new tools for plant disease diagnostic programs. Phytopathology 90:S113.

4. Wilson, A. D., and Lester, D. G. 1997. Use of an electronic-nose device for profiling headspace volatile metabolites to rapidly identify phytopathogenic microbes. Phytopathology 87:S116