Multi-Epitope Protein as a Tool of Serological Diagnostic Development for HTLV-1 and HTLV-2 Infections

Abstract

A multi-epitope protein expressed in a prokaryotic system, including epitopes of Env, Gag, and Tax proteins of both HTLV-1 and HTLV-2 was characterized for HTLV-1/2 serological screening. This tool can contribute to support the implementation of public policies to reduce HTLV-1/2 transmission in Brazil, the country with the highest absolute numbers of HTLV-1/2 infected individuals. The chimeric protein was tested in EIA using serum/plasma of HTLV-infected individuals and non-infected ones from four Brazilian states, including the North and Northeast regions (that present high prevalence of HTLV-1/2) and Southeast region (that presents intermediate prevalence rates) depicting different epidemiological context of HTLV-1/2 infection in our country. We enrolled samples from Pará (n = 114), Maranhão (n = 153), Minas Gerais (n = 225) and São Paulo (n = 59) states; they are from blood donors' candidates (Pará and Minas Gerais), pregnant women (Maranhão) and HIV+/high risk for sexually transmitted infection (STI; São Paulo). Among the HTLV-1/2 positive sera, there were co-infections with viral (HTLV-1 + HTLV-2, HIV, HCV, and HBV), bacterial (Treponema pallidum) and parasitic (Trypanosoma cruzi, Schistosma mansoni, Strongyloides stercoralis, Entamoeba coli, E. histolytica, and Endolimax nana) pathogens related to HTLV-1/2 co-morbidities that can contribute to inconclusive diagnostic results. Sera positive for HIV were included among the HTLV-1/2 negative samples. Considering both HTLV-1 and HTLV-2-infected samples from all states and different groups (blood donor candidates, pregnant women, and individuals with high risk for STI), mono or co-infected and HTLV-/HIV+, the test specificity ranged from 90.09 to 95.19% and the sensitivity from 82.41 to 92.36% with high accuracy (ROC AUC = 0.9552). This multi-epitope protein showed great potential to be used in serological screening of HTLV-1 and HTLV-2 in different platforms, even taking into account the great regional variation and different profile of HTLV-1 and HTLV-2 mono or co-infected individuals.

Keywords: HTLV-1; HTLV-2; co-infections; multi-epitope protein; serological diagnostic.

Figure 1

Western blot of purified HTLV-1/HTLV-2 multi-epitope protein. MW, Molecular weight. (1) seronegative pool 1:100/IgG anti-human 1:2,000; (2) positive HTLV-1 pool 1:100/anti-human IgG 1:2,000; (3) positive HTLV-2 pool 1:100 / anti-human IgG 1:2,000; (4) anti-histidine 1:1,500/anti-mouse IgG 1:10,000 (positive control). The black arrow indicates the molecular weight of 70 kDa, corresponding to the size of the recombinant protein. It is observed that the multi-epitope protein was detected by pool sera of individuals infected with HTLV-1 or HTLV-2 (lines 2 and 3, respectively) and by the positive control (line 3—anti-histidine). The pool of seronegative individuals did not react with the target protein (line 1).

Figure 2

Antibody reactivity of the evaluated HTLV-1, HTLV-2, HTLV, and seronegative samples using the indirect in-house ELISA based on the multi-epitope recombinant protein. (A) Scatter plots show the reactivity against the multi-epitope recombinant protein for all samples tested, which were defined as positive, indeterminate, or negative. HTLV-1 mono or co-infected x HTLV-SN or HTLV-SN/HIV (p < 0.0001), HTLV-2 mono-infected × HTLV-SN (p = 0.0005) and × HTLV-SN/HIV (p = 0.00261), HTLV-2 co-infected × HTLV-SN and HTLV-SN/HIV (p > 0.05). For HTLV-SN × HTLV-SN/HIV the difference was p = 0.0015. (B) Scatter plots comparing HTLV-1/2 groups with HTLV negative. (C) Comparison between HTLV positive × HTLV negative. Kruskal–Wallis and Mann–Whitney tests were performed in GraphPad Prism 8.0.2.

Figure 3

ROC curve analysis of Brazilian global results. HTLV: area under the curve: 0.9552 (95% CI: 0.9345–0.9758), optimal cut-off: >1.110, with 85.80 % sensitivity and 94.60% specificity at cutoff. HTLV-1: area under the curve: 0.9821 (95% CI: 0.9960–0.9982), optimal cut-off >1.110, 96.77% sensitivity, and 94.60% specificity at cutoff. HTLV-2: area under the curve: 0.8844 (95% CI: 0.8270–0.9419), optimal cut-off > 0.9150, 64.29% sensitivity and 90.23% specificity at cut-off.

Figure 4

Detection of the multi-epitope chimeric protein by human IgG from individuals infected or not by HTLV in samples from four different Brazilian states. Index values (Abs/cut off) of individuals infected with HTLV-1, HTLV-2, and HTLV and uninfected (SN) in the four states tested—Pará, Maranhão, Minas Gerais, and São Paulo. Ind.: Indeterminate status with CMIA or EIA positive and WB indeterminate pattern (Maranhão and São Paulo). Some HTLV-1 and HTLV-2 samples from Minas Gerais and São Paulo had coinfections, as described in Table 1 and include viral, parasitic and bacterial infections for Minas Gerais and HIV for all samples from São Paulo. HTLV untyped means that WB did not differentiate the infection as HTLV-1 or HTLV-2.