A field evaluation of a gE indirect ELISA for surveillance of IBR vaccinated and IBR free milk herds

INTRODUCTION Bulk milk samples have been proven to be a practical and cost-effective biological material for surveillance of many animals diseases. The application of bulk milk testing in IBR vaccinated herds has been limited for the intrinsic low sensitivity of blocking gE ELISA (1), when applied to a matrix with low IgG content. We recently developed a recombinant based gE indirect ELISA which overcome these limitations. Objective of this study was to evaluate the applicability of the method under field condition. MATERIAL AND METHODS The glycoprotein E of BoHV1 was expressed in secreted form in mammalian system, purified and coated in ELISA plates. In addition, a rapid IgG purification step was applied as upstream protocol to further increase diagnostic sensitivity in bulk milk testing. In this step, IgG are captured from milk whey by an affinity matrix and both purified and concentrated in a small elution volume. ELISA results were expressed as percent of reactivity against positive control included in each plate. Samples with a reactivity greater than the 40% of the positive control and lower the than 30% were classified as positive and negative respectively. Reactivity between the 30% and 40% was considered as doubtful. RESULTS Laboratory simulation suggested a detection limit of one positive sample out of 40 samples (theoretical prevalence of infection: 2,5 %), which seemed adequate to reveal minor seroconversion, as it could be expected in marker vaccinated herds according to EU regulation (2). Bulk milk samples were collected in three different Italian provinces (Cuneo, Torino and Trento) from positive (n=51), vaccinated (n = 114) and IBR-free (n = 55) herds. Among positive farms, 37 had a seroprevalence greater than 2.5% in lactating animals. Each samples was subjected to IgG purification/concentration and assayed in indirect ELISA. Results were compared with IBR status as detected by officially method according to the last available individual blood testing (using blocking gE ELISA), estimating the correct seroprevalence in lactating animals. When this estimation was available, ELISA test returned positive results if the seroprevalence greater than 2.96%, but tested negative if the prevalence was lower than 2.38%. This result confirmed the 2.5% of seroprevalence as theoretical limit of detection and allows to asses the sensitivity of the diagnostic process to 100% (95%CI 90.5% - 100%) in farms with a seroprevalence greater than 2.5%. On the other hand, results on gE negative farms showed a nearly perfect specificity, with no differences in term of reactivity between vaccinated and IBR-free farm samples. Indeed, in the first test session, only 1 IBR-free and 1 vaccinated farm showed positive results in ELISA. In order to better investigate the two discordant situations, we collect individual blood and milk samples and re-tested all the animals. In the IBR-free farm, the second test detected the presence of a single positive animal out of 67. The positivity was proved using both individual milk and blood sample. The positive cow had been vaccinated in 2003 with an attenuated conventional vaccine: its antibody titers decreased until 2013, testing negative at the previous gE blocking ELISA. This animal, erroneously not removed by the farmer, probably reactivated the vaccine strain with no viral excretion: this scenario was sufficient to obtain positive outcomes in ELISA test using bulk milk and can explain our results. The second discordant farm is still under investigation. To date, given these results on gE-negative samples, we can fix the specificity of the diagnostic protocol to 99.4% (95%CI 96.7%-100%), reaching the 100% in IBR-free farms (95%CI 93.4%- 100%). DISCUSSION AND CONCLUSIONS Performances of the proposed method seems to overcome the limitations of the blocking ELISA, in term of both sensitivity and specificity. As far as sensitivity in bulk milk is concerned, it has been significantly improved in comparison with conventional gE-Elisa reactions, and now it is close to the sensitivity of the most sensitive Elisa kits for total antibodies which are commercially available.