Novel approaches in the detection of Yellow Fever Virus in febrile patients visiting selected health facilities in Western Kenya

Abstract

Globally, the threat of emerging and re-emerging arboviral infections continues to be a major public health concern. The detection and diagnosis of many of these diseases remains a challenge, and it is clear that novel approaches to their identification and control must be continually developed. Clearly there is an overarching need to accurately detect priority febrile illnesses and their causative agents in Kenya. This study designed and optimized a Reverse-Transcription Isothermal Loop Mediated Assay (RT-LAMP) for yellow fever virus (YFV) assay, developed and optimized ELISA systems using anti-yellow fever virus monoclonal antibody (2D12) as well as used the developed assays to determine the seroprevalence of YFV in selected health facilities in Western Kenya during the period 2010-2012. The study design was experimental. RTLAMP was done isothermally at 62 °C using a real-time turbidimeter which allowed detection within 1 h. RT-PCR and real time RT-PCR using YFV primers was also performed. Specificity of RT-PCR, real-time RT-PCR and RT-LAMP assays was determined using RNA from other related flaviviruses (WNV, DENV1-4, JEV) as well as RVFV and CHIKV where only YFV RNA was detected. In addition, equal sensitivity was also observed when the RT-LAMP assay and the quantitative real-time RT-PCR were compared using YFV spiked human serum samples with a detection limit of 0.29 PFU. The RT-LAMP assay further showed equal detection limit sensitivity of 0.29 PFU while using two local YFV wild isolates and the vaccine strain YFV 17D. A total of 469 serum samples from febrile patients were tested for YFV antibodies using in-house IgM- xxi capture ELISA, in-house IgG indirect ELISA and focus reduction neutralization test (FRNT50). The present study did not yield any IgM ELISA positive cases an indication of absence of recent YFV infection in the area. Twenty eight samples (6%) tested positive for YFV IgG. These 28 cases detected during the study were due to either YFV vaccination or past exposure to various flaviviruses including YFV. There were five cases confirmed by FRNT50, of these, four cases were either vaccination or natural infection during YF outbreak (1992-93) or another period and one case was confirmed as a West Nile virus infection. The YFV NS1 recombinant antigen was successfully cloned and expressed in a pET 28a system. Sequencing data showed 99% homogeneity with the consensus YFV 17D strain. Confirmation of the expressed YFV NS1 was performed using a penta-his tag in a Western blot.