UBC Theses and Dissertations
Variability in an enzyme-linked, immunosorbent assay (ELISA) for Erwinia carotovora subsp. atroseptica Caron, Michel
Factors affecting a double antibody sandwich enzyme-linked immunosorbent assay (ELISA) for the detection of Erwinia carotovora subsp. atroseptica were investigated. Optimum reaction conditions for detecting known cell numbers of Eca were found to be 2.μg/ml of coating γ-globulin and 1:4-00 enzyme-γ-globulin conjugate dilution. These conditions were determined using antiserum produced against glutaraldehyde-fixed, whole bacterial cells of strain E82 of Eca (serogroup I), and polystyrene microtitration plates (Dynatech substrate plates). In spite of these optimized conditions, variability was observed between sets of data obtained under identical experimental conditions. In order to minimize or eliminate this variability, different parameters were investigated. The washing procedure was standardized by the use of a controlled pressure-washing system employing distilled water, and two 15-sec washes at 34.4-8 kPa (5 psi), with 180° rotation of the plate between each wash. Tween-20 was eliminated from the washing solution, since it interferred with the sensitivity of the assay. This effect could not be related to the age of the Tween-20 employed. Well to well variability was observed with the polystyrene microtitration plates employed but it was not exclusive to the outside rows. The pattern of distribution of the "odd" wells within a plate changed, and the number of "odd" wells decreased with time. The maximum variation from the mean also decreased with time. Addition to different wells of an extra 5% of coating y-globulin, sample, and enzyme-y-globulin conjugate individually or in different combinations, failed to reproduce the variability observed thereby eliminating pipetting errors as a source of variability. The A[sub=+.05] values were influenced by the buffer solutions employed for sample and conjugate dilution. Any given buffer had a greater effect when used for conjugate dilution. The complete buffer of phosphate buffered saline (PBS)+ 0.05% Tween-20 +2.0% polyvinylpyrrolidone+0.2% egg albumin commonly used in virus work, was found to be suitable for the Eca system although its efficiency in the presence of plant material containing bacteria remains to be evaluated. This ELISA for Eca employing optimized coating and conjugate, a standardized washing procedure and a complete buffer for samples and conjugate dilution, routinely detected 10⁵ to 10⁶ cells/ml of only serogroup I of Eca when pure cultures of both homologous and heterologous strains were tested. At concentrations >10⁷ cells/ml, strains from serogroups XVIII, XX, and XXII of subsp. atroseptica and a few strains from serogroups II, III, IV, and V of subsp. carotovora also reacted. Even at high bacterial concentration (10° cells/ml) no cross reactions were observed with Pseudomonas marginalis and Corynebacterium sepedonicum. Heat treatment of cell suspensions of serogroup I at 60 C for 3-6 min enhanced A[sub=+.05] values but the level of sensitivity was not reduced below 105 cells/ml. Cross reactions with strains of subsp. caro- tovora serogroups III and V, observed at 10 cells/ml, were reduced but not eliminated by this heat treatment. Both heat-labile and heat-stable water-soluble antigens were detected by this ELISA for Eca. The media upon which cells were grown also affected the A[sub=+.05] values but this effect was not proportional to the amount of growth observed. Based on these results it was concluded that until well to well variability is eliminated, and sensitivity increased, there will be little incentive to use the double sandwich ELISA technique with plant sap where a reduction in sensitivity is likely. At this point ELISA seems to have little potential in routine surveys for detecting latent blackleg infections in certified seed potatoes.