Comparison of immunofluorescence assay and immunomagnetic electrochemiluminescence in detection of Cryptosporidium parvum oocysts in karst water samples

Immunofluorescence assay (IFA) and immunomagnetic electrochemiluminescence (IM-ECL) were used for comparison of the percent recovery of Cryptosporidium parvum in environmental water samples obtained from a spring draining a karst basin. The monoclonal antibodies to C. parvum, isotype IgG3 were used for optimization of the IM-ECL protocol. The combination of biotinylated and TAG-labeled anti-C. parvum antibodies with the streptavidin beads gave a linear regression slope for log ECL vs. log fresh oocysts of 0.79 (from 5 to 5,000 oocysts), which indicates a constant ECL signal per oocyst. Standard curves gave a dynamic range of 5 to 5,000 oocysts/ml (fresh) and 10 to 100,000 cells/ml (4-month-old oocysts) with the maximum limit of linear detection higher than 100,000. The linear slope of 4-month-old oocysts decreased to 0.62, which indicates that ECL signal is a function of oocyst age. The experiment associated with bead storage time shows that even after 4 months of storage of the biotinylated antibodies, the complex retains the ability for binding the oocysts and generating the ECL signal. Based on the IFA results in the experiment evaluating different protocols for oocysts recovery from karst water samples, the most efficient protocol involved dispersion, followed by flotation and immunomagnetic separation (IMS) (24% recovery). The ECL results obtained in that experiment were very similar to the results obtained in the IFA method, which indicates that the IM-ECL method is accurate. Results of the IFA in the study of the prevalence of C. parvum in the groundwater showed that oocysts were present in 78% of 1 L water samples with average number of oocysts of 6.4+/-5.5 and ranged from 0 (13 samples) to 23.3 (2 samples). The ECL signal generated from these water samples ranged from 3,771 to 622 (average 1,620+/-465). However, the background value estimated in groundwater samples with low number of oocysts detected by IFA was highly variable and elevated (from 3,702 to 272, average 1,503+/-475). The background value as a result of nonspecific binding to beads by unidentified organic components in the water can inhibit or even completely mask the signal generated by oocysts. Our investigations showed that the IM-ECL method appears to be promising for the qualitative and quantitative detection of C. parvum from the environmental water; however, the method requires further development to improve sensitivity and account for background signals.