Enhanced POC assay for TB in HIV-infected children based on the ultrasensitive detection of the urinary form of the lipoarabinomannan antigen

Abstract TB remains a major source of mortality in young children in endemic areas, particularly in areas where there is a high rate of HIV infection. These children are difficult to diagnose with existing assays, and it is believed that many of the >230,000 children who died because of TB in 2017 could have been saved with better diagnostics. A highly promising biomarker for diagnosing TB is lipoarabinomannan (LAM), a major mycobacterial surface glyolipid that accumulates at different concentrations in the great majority of actively infected TB patients. Our lab has recently made major contributions towards understanding the diversity of the antigenic properties of LAM and the complexity of the humoral immune rsponse against this antigen. One surprising discovery resulting from our work is that the urinary form of TB LAM (uLAM) is antigenically distinct from the bacterially associated form (ManLAM), and we have identified novel combinations of monoclonal antibodies that allow the sensitive and specific detection of uLAM. This has recently led to the production of a new lateral flow assay (the Fujifilm SILVAMP TB LAM assay) that has a significantly increased sensitivity for uLAM over that of the commercial assay. However, despite this improvement, the sensitivity and robustness of this assay is still not sufficient to meet the WHO requirements for an optimal assay. This proposal will build on our previous work to further improve the affinities and specificities of our existing reagents, and identify new reagents with increased sensitivity and accuracy for uLAM. To better understand the nature of uLAM and provide standardized samples for future assays we will solate large volumes (~ 2L) of urine from positive patients with a range of uLAM levels, and freeze away multiple aliquots of these samples for future use. A selected set of these urines will be used to purify high concentrations of uLAM from multiple patients, and the structural and immunological properties of these antigens will be compared. We have engineered forms of existing capture and detection reagents that have improved affinities/avidities for ManLAM, and the sensitivity of these reagents will be compared to those of existing reagents and lateral flow assays against several cohorts of pediatric urine samples that meet the requirements of this RFA (20% HIV+). We will screen memory B cell populations from selected patients to identify new mabs with high affinities and/or specificities for uLAM, and select yeast display libraries generated by randomization of the key CDR regions of our lead detection reagent for variants with higher affinities and specificites than the parental antibody. The binding properties of these new mAbs will be characterized, and their efficacy in detecting uLAM in the pediatric samples will be tested. Antibody combinations with demonstrably better sensitivities than existing reagents will be transferred to our commercial collaborators at Fujifilm and other companies for incorporation and testing in their platforms. The overall goals of this study are to develop an improved LAM detection assay that meets or exceeds the Target Product Profile (TPP) proposed by the WHO for high priority biomarker-based, non-sputum-based, POC tests for TB detection.