Research Article: Modulated Zika virus NS1 conjugate offers advantages for accurate detection of Zika virus specific antibody in double antigen binding and Ig capture enzyme immunoassays

Date Published: August 2, 2019

Publisher: Public Library of Science

Author(s): Richard S. Tedder, Steve Dicks, Samreen Ijaz, Nathalia Caroline Santiago de Souza, Anderson Vincente de Paula, Flavia Levy, Raquel Medialdea-Carrera, José Eduardo Levi, Claudio S. Pannuti, Patrícia Carvalho de Sequeira, David W. G. Brown, Ines Ushiro Lumb, Giovanna Barba-Spaeth.


The accurate diagnosis and seroprevalence investigations of Zika virus (ZKV) infections remain complex due to cross reactivity with other flaviviruses. Two assay formats, both using labelled Zika virus NS1 antigen as a revealing agent (a double antigen binding assay, DABA, and an immunoglobulin Ig capture assay, G capture) were initially developed and compared with the indirect EuroimmunZ assay for the detection of anti-Zika antibody. Of 147 pre-Zika period serum samples, 39 (27%) were reactive in the EuroimmunZ or the DABA assays, 28 sera concordantly so. Such false reactivity was influenced by the serotype of Dengue virus (DV) to which individuals had been exposed to. Thus, of sera from patients undergoing secondary Dengue virus infection of known serotype, 91%, 45% and 28% of Dengue virus serotype 2, 3 and 4 respectively were reactive in one or more of the three assays. A novel method of quenching false sero-reactivity was therefore developed for the DABA and G capture assays. Initial addition of a single homologous Dengue virus serotype 3 NS1Ag quench significantly ablated false reactivities in the pre-Zika period sera. An equipotent quadrivalent quench comprising homologous Dengue virus serotypes 1 to 4 NS1Ag was shown to be optimum yet retained sensitivity for the detection of specific anti-Zika antibody. Comparing DABA and G capture assays using quenched and unquenched conjugates in comparison with EuroimmunZ early in the course of PCR-confirmed infection indicated that a significant component of the apparent early anti-ZIKA antibody response is likely to be due to a Zika virus-driven anamnestic anti-Dengue virus response. The increased specificity provided by homologous antigen quenching is likely to provide a significant improvement in sero-diagnostics and to be of clinical value.

Partial Text

Zika virus (ZKV), first described in a nonhuman primate in 1947, is named after a forested area in Uganda [1]. Its ability to infect humans was first demonstrated in 1952 as reviewed by Kindhauser and colleagues [2] with antibody to ZKV detected both by neutralisation and by animal protection experiments [3]. Zika virus is one of a number of closely related Flaviviruses, previously termed Arboviruses, principally transmitted through the bite of an infected Aedes mosquito, most commonly Aedes aegypti but which may also be transmitted directly between humans through sexual contact or vertically, from mother to the unborn baby. Although possible, no cases of transmission through transplantation of cells, tissues or organs have been reported to date; probable transmission through blood products have been reported [4–6].

After the first description of ZKV virus infection in Africa nearly 80 years ago and the sporadic numbers of human infections identified subsequently associated with mild erythematous illness, the first significant cluster of cases since then was reported from Yap Island in Micronesia in 2007 [2]. By the early part of 2016, the WHO declared that ZKV infection, associated as it was with severe neurological disorders in the new-born, was a public health emergency of international concern. A particular aspect arising from this decision was the requirement for specific and sensitive serological assays for the detection of anti-ZKV antibody. This was always going to be challenging because of the potential exposure to multiple co-circulating Flavi-viruses in endemic areas, with the known potential to elicit serological cross reactivity between these infections. Particular assay formats have individual attributes, an observation which underlines the value of taking a broader view when setting out to develop serological tests for infectious agents. This approach is exemplified by our recent experience with Ebola virus serology [13] and echoed by Balmaseda and colleagues [12]. Recognising that antibody to envelope proteins is often cross-reactive within subgroups of the Flavi-viridae family, Stettler and her colleagues [14] moved to use the non-structural protein NS1 in view of data suggesting that this protein was likely to provide a more species-specific antigen moiety for use as the solid phase in immunoassays. A similar conceptual approach is employed in the benchmark and widely used EuroimmunZ assay which uses rZKVNS1Ag in an indirect assay format. These observations informed our choice of the NS1 target and our decision to explore assay formats other than the indirect immunoassay (Fig 1).