Isolating novel antibodies against Eimeria infections of chickens using Next Generation Phage Display and conventional Phage Display

Poultry coccidiosis caused by the protozoan parasite Eimeria is a major cause of economic losses in the poultry industry. The resilience and short life cycle of the parasite coupled with high stocking densities in modern day commercial poultry establishments ensure that this parasite is an ever-present pathogen on poultry farms. Traditional methods of diagnosis such as oocyst morphology, counting oocysts per gram of faeces and lesion scoring are still heavily relied upon in spite of them being highly subjective and prone to under-diagnose subclinical disease. PCR based molecular diagnostics have been available since the late 1990’s but have not gained widespread application likely due to the relative expense and the need for specialist equipment. LAMP assays have recently been developed that can distinguish between the seven Eimeria species of chicken and they offer the advantage of being less subjective and not requiring expensive laboratory equipment. However, these assays still require template DNA to be obtained from scrapping mucosal cells of the intestines where the different Eimeria species are likely to be present and this requires some expertise knowledge. There is therefore a need to develop a cheap, rapid test that will help differentiate infecting Eimeria species and ideally that can eventually be applied in the field. This is particularly needful because of the tight economic margin prevalent in present day poultry production. Phage display is a powerful technology for isolation of diagnostic antibodies that bypasses the use of animals and links the phenotype of a displayed ligand to its genotype. The coupling of phage display and the screening power of next generation sequencing (NGS) has the potential to transform the discovery of novel antibodies against any target. It was hypothesized that NGPD will allow comparative screening of antibody binding events to identify novel antibodies to species-specific epitopes of Eimeria oocysts and such ligands could then be used to develop immunoassays to differentiate infecting species of Eimeria in chickens. Consequently, in this study, next generation phage display (NGPD) strategies were used to identify novel antibodies against whole Eimeria oocysts. Using a naïve scFv library and panning against whole oocysts in solution, 6 binders to Eimeria oocysts were identified. Whole oocysts could not be immobilised readily on solid surfaces and binding properties were established in fluorescent microscopy (FM) and a novel spin-down ELISA (SDA). The spin down assay on occasions gave different binding specificities compared to fluorescent microscopy but on the whole, the FM gave less positive results than the SDA. The inability to develop a high throughput assay for the antibodies as well as difficulties encountered in rescuing clones from polyclonal phage sub libraries were major challenges in this approach. In later panning experiments, the use of the camelid VHH library that uses a single antibody framework resolved the difficulty in rescuing phage candidates and conducting panning experiments on 96 well plates using oocyst lysates and distinct peptide regions of Eimeria oocyst surface displayed peptides had the potential to be applied in high throughput assays. For these panning experiments, the NGS approach was again used in comparison with a conventional colony picking method and the NGS data analysis involved ranking candidates using Z scores and removing non-specific binders. Subtractive phage panning was also conducted in later panning experiments to maximise the chances of isolating species-specific clones. This might have led to the removal of high affinity binders as no ELISA positive clones could be identified. The final part of this study assessed the bioinformatics analysis used in this study to try and develop strategies to identify VHH binders. The VHH NGS data was reassessed based on candidate frequency in the sequenced data and cohort analysis and candidates that were likely to be binders were highlighted. Future approaches suggested include targeting oocyst homogenates or Gam56 peptides, the use of an antibody library with a single region of diversity and the application of less stringent subtractive steps in the panning. Finally, bioinformatics approach that identifies clones that are enriched in selection (not growth) stages and either have relatively high frequencies or high species-specificity as analysed by Z score analysis should be employed