Sequence optimized diagnostic assay for Ebola virus detection

Human use and ethics

All methods were carried out in accordance with relevant guidelines and regulations, and all experimental protocols were approved by appropriate institutional and committee oversite and conducted in compliance with the United States Department of Defense, federal, and state statutes and regulations relating to the protection of human subjects, and adheres to the principles identified in the Belmont Report. USAMRIID Office of Human Use and Ethics reviewed this study and the de-identified samples and determined it to be Not Human Subject Research (HP-09-32). All samples were collected and de-identified in Sierra Leone at the Kenema Government Hospital, and the samples had indirect identifiers upon receipt. Samples were collected during the outbreak to perform emergency diagnostics and were not collected specifically for this study.

Viruses and samples

Multiple EBOV variants were used in this study including Kikwit (UCC# R4317a-T; GenBank ID MK028855.1), Luebo (UCC# R4521T), Mayinga (UCC# R3831T), and Makona (UCC# R4491T, GenBank ID MH121161.1). All viruses are maintained by the Unified Culture Collection (UCC) at the US Army Medical Research Institute of Infectious Diseases (USAMRIID). Cell culture supernatant containing virus and human serum samples were inactivated using TRIzol LS (Thermo Fisher Scientific, Waltham, MA), and total nucleic acids were isolated using the EZ1 Virus Mini Kit V 2.0 (Qiagen, Valencia, CA) and the EZ1 Advanced XL (Qiagen) according the manufacturers’ recommendations. Samples were eluted in 90 µl elution buffer and stored at − 80 °C until use. A synthetic RNA (BioSyn, Inc., Lewisville, TX) encompassing the assay target region for EBOV Beni (GenBank# MK163673), which had a mismatch in the forward primer, was used for inclusivity testing.

EBOV sequence analysis

All of the available EBOV sequences in GenBank at the start of this study (n = 1996) were aligned (settings included: gap open cost = 10.0, gap extension cost = 1.0, end gap cost free, and alignment very accurate) and analyzed (Find Binding Sites and Create Fragments tool for primers and probes) using CLC Genomics Workbench. No forward Ebo-TM primers required modification. Reverse primers and probe were modified as detailed in Fig. 1. Probe G9A was modified to be an exact match to EBOV Makona. Reverse primer G14A was modified to be an exact match to EBOV Kikwit and Makona. Reverse primer G14A, G18A was modified to be an exact match to EBOV Luebo, and reverse primer G18A incorporated the mismatch at the 3’ end of the reverse primer for EBOV Luebo. In order to assess the impact the nucleotide mismatches had on assay performance, the Ebo-TM assay was run as previously described¹⁵ with different combinations of the primer and probe.

Real-time RT-PCR

Briefly, 5 µl of extracted nucleic acid from EBOV Kikwit, Mayinga, Makona, and Luebo were assayed using the original Ebo-TM as developed or modified by having (1) the probe switched with probe G9A, (2) Ebo-TM with reverse primer G14A, and (3) Ebo-TM assay with reverse primer G14A, G18A. Assays were run on the Roche LightCycler 480 (Roche Applied Science, Indianapolis, IN) using the SuperScript One-Step RT-PCR Kit (Thermo Fisher Scientific). Cycling conditions were: 50 °C for 15 min; 95 °C for 5 min; 45 cycles (95 °C × 5 s, 60 °C × 20 s); and 40 °C × 30 s. Fluorescence readings were taken following each of the 45 cycles, and a sample was considered positive if the quantification cycle (Cq) was less than 40 cycles. The difference between the Ebo-TM assay and the modified assay (ΔCq) was determined.

Based on the sequence analysis, a new reverse primer (R2079-1, 5′-CAGTCCGGTCCCARAATRTG) and probe (p2058A-1, 5′- 6FAM-CATGTGCCRCCSCATCGCTGC-TAMRA) were designed. This probe had an additional degenerate nucleotide (the R) that was identified in several newly sequenced EBOV isolates submitted to GenBank after the analysis above was completed. The newly optimized assay was run as described above with the original forward primer. Preliminary LODs using both the Ebo-TM assay and the new EBO-TM2 assay were determined using EBOV Makona, Kikwit, Mayinga, and Luebo. Virus was serially diluted 1:10 in nuclease-free water, and the preliminary LOD was the lowest concentration in which all three test replicates were positive. The preliminary LOD was then confirmed by running 60 replicates for each virus at the preliminary LOD. If at least 58/60 replicates were not positive, 60 replicates were repeated at a higher virus concentration until at least 58/60 replicates tested positive.

Exclusivity testing was conducted using the following organisms at 100 pfu/reaction: Sudan virus [Boniface (UCC# R4142S) and Gulu (UCC# R4389T)], Reston virus H28 (UCC# R4387T), Tai Forest virus (UCC# R4371T, GenBank ID MH121167.1), Bundibugyo virus Uganda (UCC# R4386T), Marburg virus [Musoke (UCC# R4383T), Ci67(UCC# R4250T), Ravn (UCC# R4374T), and Angola (UCC# R4185T)], Lassa virus Josiah [(UCC# R4086T), Weller (UCC# R3983a-T), Macenta (UCC# R4341T), and Pinneo (UCC# R4304T)], Mobala virus (UCC# R4043T), Mopeia virus Mozambique (UCC# R3993a-T), dengue virus [serotype 1 (WestPac, UCC# R4423), 2(S16803, UCC# R4424), 3 (CH53489, UCC# R4425), and 4 (341750, UCC# R4426)], yellow fever virus (UCC# R4318T), Rift Valley fever virus (UCC# RV004a-T), West Nile virus [EG101 (UCC# R4310T) and NY99 (UCC# R4272T)], Chikungunya virus 008 (UCC# R4215T) and B (UCC# R4261T), and Crimean-Congo hemorrhagic fever virus IbAr10200 (UCC# R4401T).

Statistics

GraphPad Prism 5 was used to assess the significance of the impact nucleotide mismatches had on detection. Significance was determined using a t-test (Holm-Sidak method), and a comparison was considered significant if the difference was less than 0.05.