Abstract
Identifying the molecular mechanisms controlling the host's response to infection with Onchocerca volvulus is important to understand how the human host controls such parasitic infection. Little is known of the cellular immune response upon infection with O. volvulus. We performed a transcriptomic study using PAXgene-preserved whole blood from 30 nodule-positive individuals and 21 non-endemic controls. It was found that of the 45,042 transcripts that were mapped to the human genome, 544 were found to be upregulated and 447 to be downregulated in nodule-positive individuals (adjusted P-value < 0.05). Pathway analysis was performed on this set of differentially expressed genes, which demonstrated an impact on oxidative phosphorylation and protein translation. Upstream regulator analysis showed that the mTOR associated protein RICTOR appears to play an important role in inducing the transcriptional changes in infected individuals. Functional analysis of the genes affected by infection indicated a suppression of antibody response, Th17 immune response and proliferation of activated T lymphocytes. Multiple regression models were used to select 22 genes that could contribute significantly in the generation of a classifier to predict infection with O. volvulus. For these 22 genes, as well as for 8 reference target genes, validated RT-qPCR assays were developed and used to re-analyze the discovery sample set. These data were used to perform elastic net regularized logistic regression and a panel of 7 genes was found to be the best performing classifier. The resulting algorithm returns a value between 0 and 1, reflecting the predicted probability of being infected. A validation panel of 69 nodule-positive individuals and 5 non-endemic controls was used to validate the performance of this classifier. Based on this validation set only, a sensitivity of 94.2% and a specificity of 60.0% was obtained. When combining the discovery test set and validation set, a sensitivity of 96.0% and a specificity of 92.3% was obtained. Large-scale validation approaches will be necessary to define the intended use for this classifier. Besides the use as marker for infection in MDA efficacy surveys and epidemiological transmission studies, this classifier might also hold potential as pharmacodynamic marker in macrofilaricide clinical trials.