Experimental vaccination for onchocerciasis and the identification of early markers of protective immunity
Duprez, Jessica Anais Sybille
Onchocerciasis, caused by Onchocerca volvulus remains a major public health and socio-economic problem across the tropics, despite years of mass drug administration (MDA) with Ivermectin to reduce disease burden. Through modelling, it has been shown that elimination cannot be achieved with MDA alone and additional tools are needed, such as vaccination, which remains the most cost-effective tool for long-term disease control. The feasibility behind vaccination against O. volvulus can be demonstrated in the Litomosoides sigmodontis mouse model, which shows that vaccine induced protection can be achieved with immunisation using irradiated L3, the infective stage of L. sigmodontis and with microfilariae (Mf), the transmission stage of the parasite. There is further evidence of protective immunity in humans, with individuals living in endemic areas that show no signs of infection despite being exposed to the parasite (endemic normal). The protective efficacy of promising vaccine candidates were evaluated using an immunisation time course in the L. sigmodontis model, using either DNA plasmid or peptide vaccines. In immunisation experiments in L. sigmodontis, Mf numbers are used as a measure of protection and marks the end of an immunisation time course. However, when changes in gene expression were measured at the end of an immunisation time course, in attempts to identify gene signatures that could be used as markers of protection (correlates of protection) in the blood, no gene signatures were found to be associated with protection. This suggest that at the end of an immunisation time course, when protection is measured (change in Mf numbers), it is too late in infection to measure changes in immune pathways being triggered. Changes in gene expression were therefore measured in blood samples collected throughout an immunisation time course in the L. sigmodontis model, in order to identify the time point in an immunisation experiment which are the most indicative of protection. Two independent immunisation time courses were used, either using irradiated L3 or Mf as vaccine against L. sigmodontis, as these elicit the greatest protection. This generated a large high dimensional dataset, that was too large and complex for a differential fold-change analysis. Therefore, an analysis pipeline was created using machine learning algorithms, to detect changes in gene expression throughout the time courses to detect markers of protection. The 6 hour time point following immunisation showed the greatest change in gene expression, with the analysis pipeline identifying known pathways associated with vaccine-induced immunity. The pipeline was applied to gene expression data from human samples obtained from individuals living in endemic areas who were either infected with O. volvulus or endemic normal (naturally protected), this was to identify pathways associated with protective immunity in humans. When comparing vaccine induced immunity seen in mice and natural protective immunity in humans there was some overlap in pathways being triggered, suggesting that similar pathways are needed for protection and that if a vaccine can trigger the right pathways in mice, it is likely to be effective in humans. Overall the machine learning analysis of the gene expression data, not only shows that it is feasible to measure change in gene expression in blood during filarial infections, but that during an immunisation time course it is the early time points following immunisation that are the most predictive of vaccine efficacy (protection outcome). One of the vaccine candidates, cysteine protease inhibitor-2 (CPI), is a known immuno-modulator that inhibits MHC-II antigen presentation on antigen presenting cells such as dendritic cells (DC). This candidate has consistently been shown to induce protection if its immuno-modulatory active site was modified. In in vitro studies, it was shown that modification of the active site of CPI rescues antigen presentation in DC. This shows the importance of DC activation before the onset of infection, demonstrating the importance of triggering protective responses early in infection, and provides insight on how one of the vaccine candidates achieves protection.