© 2007 Plant Management Network.
A Comparative Analysis of Diagnostic Protocols for Detection of the Asian Soybean Rust Pathogen, Phakopsora pachyrhizi
Wayne M. Jurick II, Department of Plant Pathology, University of Florida, Gainesville 32611; Carrie L. Harmon, Southern Plant Diagnostic Network, University of Florida, Gainesville 32611; Jim Marois, Department of Plant Pathology, North Florida Research and Education Center, University of Florida, Quincy 32351; David L. Wright, Department of Agronomy, North Florida Research and Education Center, University of Florida, Quincy 32351; Kurt Lamour, Department of Entomology and Plant Pathology, University of Tennessee, Knoxville 37996; Anne Vitoreli, Department of Plant Pathology, University of Florida, Gainesville 32611; Tom Creswell, Department of Plant Pathology, North Carolina State University, Raleigh 27695; Don Hershman, Department of Plant Pathology, University of Kentucky, Princeton 42445; Consuelo Estevez, University of Puerto Rico, Juana Díaz 00795; Bob Kemerait, Department of Plant Pathology, University of Georgia, Tifton 31793; Clarissa Balbalian, Department of Entomology and Plant Pathology, Mississippi State University, Mississippi State 39762; and Philip F. Harmon, Department of Plant Pathology, University of Florida, Gainesville 32611
Jurick, W. M. II, Harmon, C. L., Marois, J., Wright, D. L., Lamour, K., Vitoreli, A., Creswell, T., Hershman, D., Estevez, C., Kemerait, B., Balbalian, C., and Harmon, P. F. 2007. A comparative analysis of diagnostic protocols for detection of the Asian soybean rust pathogen, Phakopsora pachyrhizi. Online. Plant Health Progress doi:10.1094/PHP-2007-0531-01-RS.
Plant diagnosticians routinely process soybean rust samples with one or more of the following diagnostic protocols: visual assessment using a dissecting microscope, an enzyme-linked immunosorbant assay (ELISA), and pathogen-specific PCR (conventional and real-time). Results of these three diagnostic protocols and time requirements for diagnosis were collected for samples with signs (sporulation) of soybean rust (presumed positive) and samples without sporulating rust pustules (presumed negative). Data were collected for samples processed in triplicate at six university plant disease clinics. The effects of sample treatment and storage on results of two diagnostic protocols also were examined. Visual diagnosis required the least amount of time (4 min), ELISA and PCR took approximately six-and-a-half times longer per sample (multiple samples being processed at one time) than a visual diagnosis. ELISA could not detect the pathogen in positive samples after desiccation followed by long-term storage at room temperature and following one or more autoclave treatments for 30 min. However, conventional PCR was capable of detecting P. pachyrhizi-infected plant material following all sample storage regimes.
Asian soybean rust (ASR) is caused by the filamentous, basidiomycete, plant pathogen Phakopsora pachyrhizi Sydow and was identified for the first time in the continental United States in Baton Rouge, LA in November 2004 on soybean, Glycine max (L.) Merr (9). Phakopsora pachyrhizi is known to infect a wide range of cultivated and weedy legumes including 31 species in 17 genera (3). In Florida, Pueraria lobata (Willd) Ohwi, kudzu, acts as an over-winter host for P. pachyrhizi and may serve as an important source of inoculum for neighboring soybean fields (4). Another rust on soybean is caused by a less aggressive Phakopsora species, P. meibomiae, which is geographically restricted to South and Central America and Puerto Rico and has to date not been reported in the continental US (8). Differentiation of the two soybean rust pathogens at the species level (P. pachyrhizi and P. meibomiae) is accomplished via PCR.
Soybean rust symptoms begin as discrete lesions on the lower leaf surface. Lesions expand and become visible on the upper leaf surface. Mature rust pustules are found on the lower leaf surface and are characterized by erumpent, volcano-like uredinia that contain beige to tan-colored urediniospores. Uredinia also have been observed on the stems, petioles, and pods of diseased soybeans (8). Urediniospores are ornamented and ellipsoid and may appear hyaline or yellowish-brown in color when viewed with a compound microscope (10).
Yield losses due to ASR have been reported as high as 70 to 80% where the disease is severe (1,2). Fungicide management options can be effective when products are applied preventatively (5). A nation-wide sentinel plot system was established to forewarn growers when the disease may threaten their area. In 2005 and 2006, ASR was found in sentinel plots in Florida and the Gulf Coast states and then into parts of the Midwest (7). Soybean producers and crop advisors rely on diagnostic labs to provide rapid and reliable diagnoses to help track the disease. The rate of spread and proximity to production areas directly impacts fungicide management decisions.
The objectives of this study were (i) to analyze the time requirements and consistency of results of three diagnostic methods applied to samples submitted at six university plant diagnostic laboratories, and (ii) to test the effects of sample storage conditions on the results of conventional PCR and ELISA techniques.
Diagnostic Evaluation of ASR Samples
Plant pathologists with extensive training on identification of uredinia and urediniospores associated with ASR agreed to participate at each of the six laboratories. Samples submitted by county faculty, agricultural consultants, producers, and other NPDN (National Plant Diagnostic Network) First Detectors. Samples were visually evaluated for signs of ASR via a dissecting microscope. If visual observations were consistent with ASR and sporulating pustules were present, then samples were presumed positive and processed in triplicate. A sample is defined as a leaflet of soybean, kudzu, or other legume species. Each diagnostic laboratory (University of Florida, University of Georgia, University of Kentucky, Mississippi State University, North Carolina State University, and University of Tennessee) processed 10 presumed positive and 10 presumed negative samples in triplicate using the ELISA Envirologix Qualiplate Kit TM (Envirologix, Portland, ME) according to the manufacturer’s instructions. Samples also were evaluated with conventional or real-time PCR protocols according to NPDN Standard Operating Procedures with primers specific for P. pachyrhizi and P. meibomiae (3). Results and time required for visual assessment, ELISA, and PCR were recorded.
Sample Storage Conditions for PCR and ELISA Diagnosis of P. pachyrhizi
To determine the effect of sample storage conditions on PCR and ELISA detection, P. pachyrhizi-colonized kudzu leaves were subjected to various storage and devitalization regimes and were processed in triplicate using both techniques. Sample storage/treatment conditions of ASR-colonized kudzu leaves were as follows: dried leaves stored at room temperature for one year, leaves stored in plastic bags at 4°C for 2 weeks, leaves autoclaved for 30 min (1, 2, or 3 times), and leaves frozen for 12 h at -20°C and thawed at room temperature (1, 2, or 3 times). Three ASR-infected kudzu leaves with sporulating lesions were analyzed per treatment. These experiments were conducted twice with three replications per treatment. A positive result for PCR indicated that a 330 bp amplicon was observed with UV illumination after staining with ethidium bromide, and a positive result for ELISA corresponded to a sample absorbance value of 450 nm that was 2.5 times greater than the healthy control.
Comparison of Diagnostic Protocols
The average time required to process an individual sample (batch time divided by total samples processed per batch) was approximately the same for ELISA and real-time PCR (~34 min) (Table 1). However, only 11 min were required for conventional PCR, and 4 min for visual diagnosis. The difference in time required to process real-time and conventional PCR samples can vary due to the extraction method implemented and the limited number of samples that can be processed in one real-time PCR run. For example, when 10 additional presumed positive samples were processed in triplicate (30 total) with an alternative extraction protocol (4) the total detection time was reduced to ~90 min to process all 30 samples, or 3 min per sample.
Table 1. Time requirements (min) for visual, ELISA, and PCR diagnosis of samples processed at six diagnostic labs.
* nd = not determined
The average time required to process a batch of samples using ELISA was 170 min, 220 for real-time PCR and 317 for conventional PCR. For ELISA, the time required was strongly correlated (80%) to the number of samples analyzed per run. However, for real-time PCR the time required per run was weakly correlated (34%) to the number of samples processed. This illustrates that even though there are virtually no equipment limitations on the number of samples that can be processed using ELISA, there was little gain in time efficiency per sample for runs with multiple samples.
The percentage of samples where results of the protocols were the same (percent agreement) was analyzed. Real-time or standard PCR was used to process samples. The two protocols were not compared to each other. Conventional PCR and ELISA agreed with visual observation 100% of the time. The occurrence of false positives accounts for the 11% disagreement between real-time PCR, ELISA, and visual observation. Real-time PCR results were consistent with ELISA and visual observation 89% of the time. The 11% of samples in disagreement included 15 samples (5 presumed negative in triplicate), that indicated positive for P. meibomiae with real-time PCR. These 15 samples were sent for federal testing (USDA) and all were determined to be negative. Additional false positive samples that make up the 11% in disagreement included 6 soybean samples that were processed (two presumed negative in triplicate) at the University of Florida from a stand of soybeans where ASR had been detected. These 6 samples possessed chlorotic flecking, but no erumpent uredinia were present. Sample results were negative for ELISA, but results were positive for real-time PCR. Although these results are false positives for this study, they are not surprising since real-time PCR has been shown to detect latent infection and is routinely used to detect random spore deposition from air and rain samplers (6,11).
Evaluation of Sample Treatment Regimes on Detection of P. pachyrhizi
None of the conditions examined precluded the detection of P. pachyrhizi using PCR (Table 2). However, ELISA did not detect P. pachyrhizi after long-term dry storage and did not detect from samples that were autoclaved 1, 2, or 3 times for 30 min. The viability of the pathogen in samples after each treatment was not evaluated; however, distinct lesions and uredinia remained visible after all treatments.
Table 2. The effect of sample storage conditions on detection of P.
Dolichus samples with a rust were collected in Puerto Rico and were processed using conventional PCR, real-time PCR, and ELISA. Samples processed with P. meibomiae primers and conventional PCR produced a diagnostic 330 bp amplicon. No product was observed when the samples were processed with P. pachyrhizi-specific primers. Real-time PCR results were positive for P. meibomiae as well. All P. meibomiae samples returned positive results with the Envirologix QualiPlate ELISA.
With proper training, diagnosticians are proficient at recognizing soybean rust after observation of pustules and spores using a dissecting microscope. PCR and real-time PCR are useful tools for confirmation of the soybean rust pathogens at the species level, especially in Florida and the Gulf Coast where P. meibomiae is most likely to occur. PCR is also useful for early detection before the development of soybean rust symptoms currently used in ASR diagnosis (6). The ELISA kit that was evaluated is not species-specific and requires too much time to be used for field application. We found that the ELISA kit was not as sensitive as PCR when detecting ASR on samples that are stored or transported under sub-optimal conditions (i.e., conditions analogous to ones examined in this report) and also did not detect latent infections. Therefore, negative ELISA results of symptomatic tissue should be further investigated using PCR where symptoms and signs are evident and or transport/storage of samples is in question. However, if an ELISA kit were developed that could be read in about 15 min or less, species-specificity would not be necessary in order to be useful as a rapid confirmation of field observations.
The authors acknowledge with gratitude collaborators at participating institutions who processed samples and collected data for this study. Funding for this project was provided by USDA-CSREES.
1. Bonde, M. R., Melching, J. S., and Bromfield, K. R. 1976. Histology of the susceptible pathogen relationship between Glycine max and Phakopsora pachyrhizi, the cause of soybean rust. Phytopathology 66:1290-1294.
2. Bromfield, K. R. 1984. Soybean rust monograph 11. The American Phytopathological Society, St. Paul, MN.
3. Frederick, R. D., Snyder, C., Petersen, G. L., and Bonde, M. R. 2002. Polymerase chain reaction assays for the detection and discrimination of the soybean rust pathogens Phakopsora pachyrhizi and P. meibomiae. Phytopathology 92:217-227.
4. Harmon, P. F., Momol, M. T., Marois, J. J., Dankers, H., and Harmon, C. L. 2005. Asian soybean rust caused by Phakopsora pachyrhizi on soybean and kudzu in Florida. Online. Plant Health Progress doi:10.1094/PHP-2005-0613-01-RS.
5. Kemerait, R. C., Jost, P. H., Sconyers, L. E., Kichler, J., and Clark, J. 2005. Assessment of select fungicides for management of Asian soybean rust. Online. Soybean Res. Ext. Rep., Univ. of Georgia.
6. Lamour, K. H., Finley, L., Snover-Clift, K. L., Stack, J. P., Pierzynski, J., Hammerschmidt, R., Jacobs, J. L., Byrne, J. M., Harmon, P. F., Vitoreli, A. M., Wisler, G. C., Harmon, C. L., Levy, L., Zeller, K. A., Stone, C. L., Luster, D. G., and Frederick, R. D. 2006. Early detection of Asian soybean rust using PCR. Online. Plant Health Progress doi:10.1094/PHP-2006-0524-01-RS.
7. Marois, J. J., Wright, D., Harmon, C. L., and Walker, J. 2005. Florida soybean sentinel plots for 2005. Online. Proc. of the Nat'l Soybean Rust Symp., Nov. 14-16, 2005, Nashville, TN. Plant Management Network.
9. Schneider, R. W., Hollier, C. A., Whitham, H. K., Palm, M. E., McEmy, J. M., Hernandez, J. R., Levy, L., and DeVries-Patterson, R. 2005. First report of soybean rust caused by Phakopsora pachyrhizi in the continental United States. Plant Dis. 89:774.
10. Sinclair, J. B. 1999. Soybean rust. Pages 25-26 in: Compendium of Soybean Diseases, 4th Ed. American Phytopathological Society, St. Paul, MN.
11. Steinlage, T., Miles, M. R., and Hartman, G. L. 2005. Detection and quantification of soybean rust (Phakopsora pachyrhizi) spores by quantitative real-time PCR. Online. Proc. of the Nat'l Soybean Rust Symp., Nov. 14-16, 2005, Nashville, TN. Plant Management Network.