Patent Application
20200215180
The invention relates diagnostic systems comprising protein variants of the wild-type E proteins of Flaviviruses (e.g., a dengue or Zika virus) for use in diagnosis of Flavivirus infection as well as to compositions, tests devices, kits, kit-of-parts, methods and uses relating thereto, in particular for specific detection of Flavivirus antibody, diagnosis of Flavivirus infection and investigation of exposure to Flavivirus.
The Flaviviridae are a family of positive, single-stranded, enveloped RNA viruses. They are found in arthropods, (primarily ticks and mosquitoes), and can infect humans. Members of this family belong to a single genus, Flavivirus, and cause widespread morbidity and mortality throughout the world. Some of the mosquito-transmitted viruses include: Dengue Fever, Zika virus, Yellow Fever, Japanese encephalitis and West Nile viruses. Other Flaviviruses are transmitted by ticks and are responsible of encephalitis and hemorrhagic diseases: Tick-borne Encephalitis (TBE), Kyasanur Forest Disease (KFD) and Alkhurma disease, and Omsk hemorrhagic fever.
Flaviviruses are small spherical virions encoding ten viral proteins: three structural (capsid, precursor membrane/membrane, and envelope (E)) and seven nonstructural proteins. The E protein has important roles in viral attachment to cells, fusion with endosomal compartments, and modulating host immune responses. The ectodomain of the virus E protein folds into three structurally distinct domains (DI, DII, and DIII) forming head-to-tail homodimers on the surface of the virion. DI is the central domain that organizes the entire E protein structure. DII is formed from two extended loops projecting from DI and lies in a pocket at the DI and DIII interface of the adjacent E protein in the dimer. At the distal end of DII is a glycine-rich, hydrophobic sequence called the fusion loop, which encompasses residues 98-110, and is highly conserved among flaviviruses. This region has been implicated in the pH-dependent type II fusion event; during this process it becomes exposed and reoriented outward, making it available for membrane contact. DIII forms a seven-stranded Ig-like fold, is the most membrane distal domain in the mature virion, and has been suggested to be involved in receptor binding. A stem region links the ectodomain to a two-helix C-terminal transmembrane anchor that is important for virion assembly and fusion.
Dengue disease is a mosquito-borne viral infection caused by dengue virus (DENV), one of the most important human pathogens worldwide. The infection produces a systemic disease with a broad spectrum of outcomes, ranging from non-symptomatic/mild febrile illness (Dengue Fever, DF) to severe plasma leakage and haemorrhagic manifestations (Dengue Haemorrhagic Fever, DHF) that can further evolve into potentially fatal conditions (Dengue Shock Syndrome, DSS). DENV is spread by Aedes spp. mosquitoes and is widely distributed throughout the tropical and subtropical regions of the world. About 3 billion people, in over 100 countries, are estimated to be at risk of infection, with over 300 million infections, 500,000 episodes of DHF manifestations and 20,000 deaths reported each year. The spread and impact of Dengue disease has led the World Health Organization to classify it as the “most important mosquito-borne viral disease in the world”.
Four different serotypes of dengue viruses (DENV1, DENV2, DENV3 and DENV4) have been identified to date; each serotype is pathogenic in humans. Infection with any one serotype induces lifelong immunity against that specific serotype, with only transient cross-protection against the three other serotypes. Severe manifestations of dengue infection are associated with secondary infections involving different viral serotypes; this happens through a mechanism known as antibody-dependent enhancement of infection (ADE). In ADE, recognition of viral particles by cross-reacting, but weakly or non-neutralising antibodies, leads to an increased Fc receptor-mediated uptake of immature or incompletely neutralised viruses by monocytes, macrophages and dendritic cells (the primary targets of dengue virus infections in humans) resulting in increased infectivity and deterioration of the patient's clinical condition. ADE is a critical consideration in dengue vaccine development, because an immunogen that does not elicit fully-neutralising antibodies to all four serotypes may contribute to disease, rather than prevent infection. Given the lack of efficient treatment against the infection and the risk to human health, there is a need to develop an efficient vaccine that provides a protective response without the potential to cause antibody-dependent enhancement.
One dengue vaccine has been licensed, Dengvaxia® (CYD-TDV), developed by Sanofi Pasteur. Approximately five additional dengue vaccine candidates are in clinical development, with two candidates (developed by NIH/Butantan and Takeda) which entered Phase III clinical trials in 2016.
In clinical trials, the Dengvaxia® vaccine was found to increase risk of hospitalization due to dengue haemorrhagic fever (the very disease it is meant to prevent) in young children (<5 years). As a result, Dengvaxia® vaccine has a limited license, i.e., only for persons of 9 years of age and above. Given the antigenic cross-reactivity of Zika and dengue, there is concern that vaccination with Dengvaxia® vaccine and other dengue vaccines under development may promote ADE of Zika virus, increasing the incidence of Guillain-Barré syndrome in adults and microcephaly in infants, and that vaccines in development against Zika may likewise increase risk of dengue haemorrhagic fever, as does Dengvaxia in some subjects.
Zika virus is a mosquito-borne flavivirus that was first identified in Uganda in 1947 in monkeys, it was later identified in humans in 1952 in Uganda and the United Republic of Tanzania. Outbreaks of Zika virus disease have been recorded in Africa, the Americas, Asia and the Pacific. From the 1960s to 1980s, human infections were found across Africa and Asia, typically accompanied by mild illness. The symptoms are similar to infections such as dengue, and include fever, skin rashes, conjunctivitis, muscle and joint pain, malaise, and headache. These symptoms are usually mild and last for 2-7 days. However, Zika virus infection may cause complications in some subjects. Zika virus infection during pregnancy has been recognised as a cause of congenital brain abnormalities, including microcephaly. Zika virus is a trigger of Guillain-Barré syndrome. Links between Zika virus and a range of neurological disorders are being investigated.
Sanofi reported in 2016 its collaboration with the Walter Reed Army Institute of Research (WRAIR) in the United States and Fiocruz public health center in Brazil to develop a Zika vaccine and reported in 2016 that immunization with a plasmid DNA vaccine or a purified inactivated virus vaccine provided complete protection in susceptible mice against challenge with a strain of Zika virus involved in an outbreak in northeast Brazil (Larocca et al., 2016 Nature 536, 474-478 (25 Aug. 2016)
However, plasmid DNA vaccination in man requires ‘gene gun’ or similar technology (e.g., electroporation) for delivery and this approach is not considered to provide a global solution to the problems of dengue and Zika. Also, both the DNA vaccine and inactivated virus vaccine approaches in development contain dengue-Zika cross-reactive epitopes implicated in the causation of ADE.
After infection, or vaccination, the body's immune system produces neutralizing antibodies that bind to the surface proteins of a virus to block infection. Antibody-dependent enhancement (ADE) occurs when antibodies elicited by one virus can bind to, but do not block (neutralise) the infection of a similar virus.
ADE is most commonly observed for dengue virus. The 4 known serotypes of dengue virus have distinct, but related surface proteins. Infection with a first dengue virus serotype typically results in mild, or no, symptoms in the infected subject. If the subject is infected subsequently with a second dengue serotype, the immune system will produce antibodies to the first serotype that bind to the second serotype of virus, but will not always block infection and which have the potential to cause ADE. As a result there is antibody-mediated uptake of virus into cells that dengue virus does not normally infect (i.e., cells having receptors for the ‘tail’ or Fc region of the antibody). This can result in a more severe form of disease such as dengue hemorrhagic fever or dengue shock syndrome. Only young infants develop dengue haemorrhagic fever upon a first exposure to dengue, as a result of transplacentally transmitted maternal anti-dengue antibodies. As such, antibodies are equal partners with virus in (severe) disease causation in adults and infants alike.
Dengue virus antibodies not only promote ADE of other dengue virus serotypes, but also enhance Zika virus infection. Dejnirattisai et al., (2016) Nature Immunology 17, 1102-1108. “Dengue virus sero-cross-reactivity drives antibody-dependent enhancement of infection with Zika virus”. Dejnirattisai et al. tested the effect of dengue neutralizing antibodies or serum from dengue virus patients on Zika virus in cell culture. In the absence of antibody, Zika virus poorly infected the cells, but when Zika virus was incubated with dengue serum or neutralizing antibodies, Zika virus robustly infected these cells, indicating the operation of ADE. The physiological relevance of this finding requires confirmation in epidemiological studies, but these findings pose an obvious risk for current vaccine approaches. To date no satisfactory solution to this problem has been conceived or advocated.
While vaccines in this field may transpire to have net benefit on a population basis, on an individual basis the picture is different. In some subjects, tragically, preventing one disease may increase the severity or risk of mortality from another. Paul L M et al. Clinical & Translational Immunology (2016) 5, e117 “Dengue virus antibodies enhance Zika virus infection” have reported that:
Dengue virus antibodies can promote ADE of Zika virus. Zika virus antibodies can promote ADE of dengue virus. Thus, immunization against Zika virus could increase the incidence of dengue hemorrhagic fever or dengue shock syndrome, or foster the development of these conditions in individuals that would not otherwise have developed them, but for immunisation. Given the interval between infections, which can be several years, it will be years before post-marketing surveillance studies are able to inform if, and to what extent, new vaccines predispose to severe dengue disease (haemorrhagic fever, shock syndrome) or severe Zika sequelae, such as Guillain-Barré syndrome or microcephaly.
Accordingly, there is a clear need for vaccine approaches that are designed purposefully to avoid the problem of antibody-dependent enhancement.
WO2016012800 discloses identification and characterisation of cross-reactive neutralising antibodies obtained from patients infected with dengue virus. The acute human antibody response was found to be focused on two major epitopes; a known epitope on the fusion loop (FL FLE), and a second epitope, said to be novel, which was found on intact virions or dimers of envelope protein and which encompassed areas of domains I, II and III. Antibodies reactive with the second epitope, the Envelope Dimer Epitope, or EDE, were reported to fully neutralise virus made in both insect and primary human cells in the low picomolar range. A subunit vaccine comprising a stabilized soluble protein E dimer was therefore proposed as a dengue vaccine. WO2016012800 discloses that a dengue virus envelope glycoprotein E ectodomain (sE; soluble envelope polypeptide/glycoprotein) refers to the 1-395 amino acid fragment of the envelope glycoprotein E of the dengue virus serotypes 1, 2 and 4, and to the 1-393 amino acid fragment of the envelope glycoprotein E of the dengue virus serotype 3. WO2016012800 described the EDE as a stabilised dimer of sE, selected from DENV-1 sE, DENV-2 sE, DENV-3 sE, DENV-4 sE and mutant sE thereof having at least one mutation (substitution) selected among H27F, H27W, L107C, F108C, H244F, H244W, S255C, A259C, T/S262C, T/A265C, L278F, L292F, L294N, A313C (S313C in DEN3) and T315C, which mutations are considered to contribute to increased stability in the dimer configuration. It is disclosed that mutant sE thereof may further comprise at least one mutation (substitution) selected from Q227N, E174N and D329N; preferably all three mutations Q227N, E174N and D329N, which mutations are said to mask non-appropriate immunogenic regions and allow the stabilized recombinant sE dimer of the invention to preferentially elicit neutralizing antibodies directed to all four dengue virus serotypes.
The sE dimer mutations described are said not to interfere with immunogenicity but to provide a higher dimer affinity, by including cysteine mutations at the dimer contacts to provide stabilization by cross-links, and/or by introduction of new glycosylation sites to allow chemical cross-linking between adjacent sugars on the dimer by click chemistry, and/or by substitution of at least one amino acid residue in the amino acid sequence of at least one sE monomer with at least one bulky side chain amino acid to allow forming cavities at the dimer interface or in domain 1 (D1)/domain 3 (D3) linker of each monomer.
WO2016012800 discloses that the envelope protein may be engineered such that an improved EDE is generated, an EDE which is incapable of being recognised or raising anti-fusion loop (anti-FL) antibodies was considered to be an improved EDE. It is disclosed that such improvement may be accomplished by one or more mutations, deletions or insertions in the envelope protein, by generating a hybrid protein wherein the specific epitope (without any antigens which would raise anti-FL antibodies) is fused to a scaffold protein, or by engineering the envelope protein by modifying the internal surface of the dimer (projecting to the inside of the virus) with sugars to make it less immunogenic by adding N or O linked glycan sequences.
Roby et al., (2013, 2014) describe an approach to development of a vaccine candidates for West Nile virus by introduction of large internal deletions within the capsid (C) gene of flavivirus genomes to generate replication-competent RNAs that are unable to be packaged into virions, yet maintain secretion of highly immunogenic subviral particles (SVPs) without generating infectious virus. Such pseudoinfectious C-deleted vaccines are able to replicate and secrete large amounts of non-infectious immunogenic subviral particles (SVPs) from transfected cells and thus are said to offer the combined benefit of the safety of noninfectious inactivated or subunit vaccines with the robust immune response generated by the replication of live vaccines.
Roby et al., (2013) generated a construct, pKUNdC/C (KUNdC18-100/CMV-C), with C-deleted CMV-promoter driven cDNA of West Nile virus Kunjin (KUNV) in which alpha helices 1, 2, and 4 were removed in two separate segments and the hydrophilic alpha helix 3 was maintained. In pKUNdC/C C-deleted WNV cDNA was placed under the control of one copy of the cytomegalovirus (CMV) promoter and the C gene was placed under the control of a second copy of the CMV promoter in the same plasmid DNA. The conservation of the larger cytosolic moiety (alpha helix 3) led to a significant improvement in SVP secretion compared to that of constructs with deletions of all alpha helices of C and dC44-59. Additional improvements to SVP secretion were also observed upon the incorporation of an Asn-linked glycosylation motif at N154 of the E protein, a feature of many circulating strains of WNV and recent isolates of KUNV, corresponding to an NYS motif at amino acids 154 to 156 of the E protein. pKUNdC/C was shown to generate single-round infectious particles (SRIPs) capable of delivering self-replicating C-deleted RNA producing SVPs to surrounding cells. However, the amounts of both SRIPs and SVPs produced from pKUNdC/C DNA were relatively low.
Roby et al., (2014) reported production of a second generation constructs with C-deleted cDNA of West Nile virus Kunjin (KUNV) in which the CMV promoter was replaced by a more powerful elongation factor EF1a promoter and different forms of C were used to attempt to increase SRIP production by optimizing trans-C expression. A construct containing an elongation factor EF1a promoter encoding an extended form of C was demonstrated to produce the highest titres of SRIPs and was immunogenic in mice. SRIP and SVP titres were further improved via incorporation of the N154 glycosylation motif in the envelope protein (corresponding to an NYS motif at amino acids 154 to 156 of the E protein) which enhanced secretion of SVPs.
Davis et al., (2014) investigated the ability of West Nile virus (WNV) to infect CD209-expressing cells. Mammalian cell-derived West Nile virus preferentially infects cells expressing the C-type lectin CD209L but not cells expressing CD209; by contrast, Dengue virus (DENV) infection is enhanced in cells expressing either attachment factor. DENV and WNV virions have very similar structures. Their surfaces consist of a regular array of 180 envelope (E) protein subunits arranged in an icosahedral lattice (36). The small membrane (M) protein, generated following furin-mediated processing of pre-membrane protein (prM), is also present on the virion surface but is mostly buried in the viral membrane. The major structural differences between DENV and WNV virions stem from the number and location of N-linked glycosylation sites in the DENV viral E proteins. Most DENV isolates contain glycosylation sites at residues 67 and 153, although the site at 153 may not always be utilized; WNV E proteins only contain an N-linked glycan at asparagine 154, although this is absent in many virus strains. The presence of N-glycosylation on the WNV E protein has been linked in some studies to increased neuroinvasiveness in mice and to altered cellular tropism in vitro. Davis et al. introduced a glycosylation site at position 67 into West Nile virus E. Reporter virus particles pseudotyped with this E protein infected cells using either CD209 or CD209L. Glycosylation sites were introduced at several other positions. The WNV strain NY99 prM-E expression plasmid pCBWN and a derivative of this plasmid lacking the N-linked glycosylation site at E protein residue 154 (NY99-N154Q) were used as templates for the introduction of novel N-linked glycosylation sites into the WNV E protein by site-directed mutagenesis. The following amino acid changes were introduced into NY99-N154Q: (i) Ala-54 to Thr (A54T) adds an N-linked glycosylation site at Asn-52; (ii) D67N adds a site at Asn-67; (iii) K84T adds a site at Asn-82; (iv) A173N and P174G (AP173NG) add a site at Asn-173; (v) Glu-182 to NGS (E182NGS) adds a site at Asn-182 by mutating Glu-182 to Asn and inserting two amino acids (Gly-Ser) to complete the sequon; (vi) S230N and V232T (STV230NTT) add a site at Asn-230; (vii) V279T adds a site at Asn-277; (viii) T301N and G303S (TYG301NYS) add a site at Asn-301; (ix) T330N adds a site at Asn-330; (x) K370T adds a site at Asn-368; (xi) G389N and Q391T (GEQ389NET) add a site at Asn-389. All sites allowed CD209Lmediated infection, but only a subset promoted CD209 use. As seen for other viruses, mannose-rich glycans on West Nile virus were required for its interactions with CD209, however, mannose-rich glycans were not required for CD209Lmediated infection. Complex glycans, particularly N-acetylglucosamine-terminated structures, were able to mediate reporter virus particle interactions with CD209L. Davis et al. proposed that that CD209L recognizes glycosylated flaviviruses with broad specificity, whereas CD209 is selective for flaviviruses bearing mannose-rich glycans and thus that the location of the N-linked glycosylation sites on a virion determines the types of glycans incorporated, thus controlling viral tropism for CD209-expressing cells.
The Zika epidemic is a global problem with profound consequences for nations and families that will take decades to unfold, with an estimated 2.3 billion people at high or very high risk of infection (Alaniz, Bacigalupo, & Cattan, 2017). While Zika is asymptomatic in 80% of cases, it can give rise to serious conditions such as Guillain-Barre' syndrome and acute disseminated encephalomyelitis in adults (Medina & Medina-Montoya, 2017), as well as microcephaly and other abnormalities of infants born to women infected during pregnancy (i.e., the Zika congenital syndrome ZCS (Lucey, Cummins, & Sholts, 2017)). While there is an urgent need for a new vaccine against Zika, there is a similar urgency and an ongoing need to accurately monitor, map and contain outbreaks of Zika—which is difficult to distinguish, in terms of its clinical presentation and symptoms, from various other infections—particularly dengue. While an ongoing infection with Zika, dengue or other flaviviruses can be accurately diagnosed by measurement of virus by reverse-transcriptase quantitative PCR, or similar methodologies that detect the viral genome in body fluids (plasma, serum, urine), the viraemic phase of these diseases is typically short, in the case of Zika lasting only a few days (Hofer, 2016), and can easily be missed. Also, nucleic-acid based tests using PCR or similar methodology are relatively expensive and not well-suited to deployment in resource-limited countries where epidemics may initiate undetected, hampering the targeted deployment vector-control measures to control mosquito populations. For these reasons there is a particular need for serological tests (which measure antibodies), that can determine if a person has been infected with Zika (or dengue), particularly serological tests, such as point-of-care tests, that can be executed without the need for centralised laboratory facilities.
In addition to the value of serodiagnostics in support of vector control campaigns for both Zika and dengue, there are additional needs for reliable serodiagnostics relating to the safe use of vaccines and the rapid mapping of ongoing Zika outbreaks for the effective and appropriate deployment of new vaccines in clinical trials, and once licensed. However, there is a problem in distinguishing, reliably, between Zika and other flavivirus infections based on existing serodiagnostic tests which are particularly prone to false positives due to antibody cross-reactivity. Thus, Zika is closely related to dengue (another flavivirus infection, spread by the same mosquito species), having about 50% sequence identity with dengue and a remarkably similar 3D structure, juxtaposition and topography of the immunogenic envelope proteins of the virion surface (Sirohi et al., 2016), explaining the high degree of antibody cross-reactivity between these viruses (Chang et al., 2017) (Priyamvada et al., n.d.) (Dejnirattisai et al., 2016).
By virtue of its cross-reactivity with dengue, Zika is part of a complex ecosystem of virus interactions where infection of a human subject with one flavivirus can give rise to antibodies that influence the course of other flavivirus infections both positively and negatively (Halstead, 2014). This phenomenon is well-established for dengue infection, where a second episode of infection (necessarily, with a different serotype of dengue) may give rise to severe disease (including dengue shock syndrome and dengue haemorrhagic fever) (Guzman, Alvarez, & Halstead, 2013). Dengue virus is responsible for 390 million infections annually, and is capable of causing life-threatening ‘severe dengue’ including haemorrhagic fever and shock syndrome. According to the WHO, “An estimated 500 000 people with severe dengue require hospitalization each year, and about 2.5% of those affected die.”. The mechanism of severe dengue upon secondary infection, though long suspected, has recently been formally demonstrated to be attributable to antibody-dependent enhancement (ADE) of disease (Katzelnick et al., 2017). Likewise dengue vaccination, with the recently licensed DengVaxia™ vaccine can also have this undesirable effect in dengue naïve subjects, predisposing them to severe dengue by in effect acting as a silent primary infection (Ferguson et al., 2016; Flasche et al., 2016; Hadinegoro et al., 2015). In order to minimise the risk of priming for severe dengue, the dengue vaccine is licensed only for persons 9 years of age and above, and in territories with high endemicity for dengue. Clearly, however, these measures cannot be expected to eliminate the risk entirely (Halstead, 2017a), such that a serological diagnostic test capable of reliably distinguishing prior dengue exposure from prior Zika exposure could allow safer deployment of the DengVaxia vaccine and potentially other flavivirus vaccines in development.
Due to the cross-reactivity of Zika with dengue, and the phenomenon of antibody-dependent enhancement, it is rationally anticipated that Zika infection, and potentially also Zika vaccination (when vaccines become available), will prime dengue-naïve subjects for severe dengue (Russell, 2016; Screaton, Mongkolsapaya, Yacoub, & Roberts, 2015; Willis & Hensley, 2017); and (conversely) it is anticipated that antibodies generated by dengue infection or vaccination will likewise be capable of ADE of Zika infection (Dejnirattisai et al., 2016; Paul et al., 2016; Screaton et al., 2015), as would be predicted by Zika's notional status as a fifth serotype of dengue. Although these concerns await formal confirmation in human epidemiological studies (Halstead, 2017b; 2017a) (Sariol, Nogueira, & Vasilakis, 2017), as did (until recently) the role of antibody dependent enhancement in causation of severe dengue (Katzelnick et al., 2017), there are substantial grounds for concern.
A reliable point-of-care diagnostic could be applied to travellers from non-dengue endemic countries to dengue-endemic countries allowing them to be vaccinated safely, without risk of predisposing them to severe dengue, extending the utility of the DengVaxia vaccine (and likely other dengue and Zika vaccines that will be licensed) to traveller populations, who are predominantly dengue naïve. The consequences of prior Zika virus infection for the DengVaxia vaccine have not yet been established, but it may reasonably be expected to have an influence, positively or negatively, on the safety and efficacy of vaccination with the DengVaxia vaccine, and other vaccines that may be licensed for dengue.
Vaccine resources are often in short supply, serodiagnostics, especially point of care diagnostics, have the potential to allow targeted distribution to outbreak areas. Also, such diagnostics can be used to inform both the effectiveness and safety of development/deployment of novel vaccines. For example antibodies (resulting from prior infections with related viruses) already present in the blood of patients experiencing Zika virus infection may influence the course of the disease or the result of Zika vaccination (e.g. whether the naturally-encountered Zika virus crosses the placenta and damages the foetus, or whether live-attenuated vaccine strains might do the same thing). Current diagnostic practice is not adequate to properly enable the intelligent deployment of vaccines because it is based in central diagnostic labs that require the transport of blood samples from diverse, sometimes remote, areas of endemic countries, which is expensive and inefficient, and which may not be possible in many instances due to lack of the necessary refrigerated transport infrastructure.
What is needed to enable effective monitoring and safe development of Zika and dengue vaccines is a cheap ‘point-of-care’ diagnostic test that can distinguish reliably between them, i.e., a test that can be run without the need for clean water, electricity or equipment—e.g., like a do-it-yourself (DIY) pregnancy test, that can be operated in the home or in a Doctor's surgery or hospital clinic. Such test will help define which subjects are eligible to receive Zika vaccination (e.g., futile to immunise a Zika immune subject), maximising the safety and effectiveness of its deployment. It will also help define whether a subject has responded adequately to the vaccine (i.e. reached a ‘to be determined’ protective level of Zika-specific or dengue-specific antibodies, e.g., by detecting neutralising antibodies) or whether they may require a further dose (a factor that may vary in differing endemic territories, depending on prior exposure to related viruses).
To date, the ‘gold standard’ with respect to specificity for the serodiagnosis of dengue, Zika and other flaviviruses is the PRNT test (plaque reduction neutralisation test), wherein, in a laboratory centre, susceptible cells are infected in the presence of various concentrations of a test serum and a 50% inhibition value reported. However, this test is prone to false positives even with diverse flaviviruses. Thus Houghton-Trivino et al. found that 16/20 dengue infected subjects had high yellow fever neutralisation titres which were not attributable to yellow fever vaccination or infection (“Dengue-yellow fever sera cross-reactivity; challenges for diagnosis.—PubMed—NCBI,” 2017). Serodiagnostic tests will face increasing challenges of cross-reactivity as new flavivirus pandemics emerge in the future (Smith, 2016), as did Zika recently, unexpectedly. Notably, there is particular concern at the time of writing about resurgent yellow fever in Brazil and Africa (Mir et al., 2017), and further concern about the introduction of yellow fever to China from Africa (Ling et al., 2016). The presence of yellow fever epidemics will further complicate serodiagnosis using tests that do not accurately distinguish between yellow fever and other flaviviruses.
Specific diagnosis of Flavivirus infections using current serological testing is complicated by the cross-reactivity between antibodies against other clinically-relevant flaviviruses. Cross-reactivity is particularly problematic in areas where different flaviviruses co-circulate or in populations that have been immunized with vaccines to Flaviviruses. The majority of cross-reactive antibodies are raised against the immunodominant flavivirus envelope (E) protein target a conserved epitope in the fusion loop at the distal end of domain II.
There is a need for a diagnostic approach that can differentiate between closely-related Flaviviruses, to assess if an individual is seronegative and thus has not been exposed to dengue or Zika, or if an individual is seropositive and has been exposed to Zika and/or dengue and for those who are seropositive, to distinguish to which of Zika and/or the four dengue serotypes the individual has been exposed. There is a need for a diagnostic approach that can be used to select subjects for immunization, or assess seroconversion to determine if immunization has raised a protective immune response against dengue or Zika. There is thus a need for diagnostic approaches that enable interrogation of the immune response to distinguish antibodies against the dengue virus serotypes and against Zika virus.
The invention provides:
1. An isolated recombinant analogue of a flavivirus E-protein comprising an analogue of a flavivirus E-protein fusion loop, wherein the analogue of the flavivirus E-protein fusion loop comprises at least one glycosylation site for an N-linked glycan that is not present in a natural flavivirus E-protein fusion loop sequence, wherein the at least one glycosylation site is an N-linked glycosylation sequon (Asn-X-Ser/Thr) and the Asn (N) residue of the sequon occupies any of positions 98-110 (DRGWGNGCGLFGK) of the natural flavivirus E-protein fusion loop amino acid sequence, wherein X is any amino acid residue except proline and Ser/Thr denotes a serine or threonine residue, for use in an in vitro method for diagnosis of flavivirus infection and/or to investigate exposure to flavivirus.
2. An isolated recombinant analogue of a flavivirus E-protein according to clause 1, wherein the analogue of the flavivirus E-protein fusion loop comprises two glycosylation sites that are not present in a natural flavivirus E-protein fusion loop.
3. An isolated recombinant analogue of a flavivirus E-protein of any preceding clause which is glycosylated with a glycan at one or at both of the introduced glycosylation sites in the analogue of the flavivirus E-protein fusion loop.
4. An isolated recombinant analogue of a flavivirus E-protein of clause 2 or clause 3 wherein the glycan is an N-linked glycan.
5. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, comprising an N-linked glycosylation sequon (Asn-X-Ser/Thr) such that an Asn (N) residue of the sequon occupies any of positions 98-101 and/or 106-110.
6. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, wherein X is any of the following 13 amino acid residues Gly, His, Asn, Gln, Tyr, Val, Ala, Met, Ile, Lys, Arg, Thr or Ser.
7. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, wherein the flavivirus E-protein is a dengue virus E-protein and the Asn (N) residue of a sequon occupies position 101, 108 or both 101 and 108 of the amino acid sequence of the flavivirus E-protein fusion loop or the flavivirus E-protein is a Zika E-protein and the Asn (N) residue of a sequon occupies position 100 of the amino acid sequence of the flavivirus E-protein fusion loop.
8. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, wherein the flavivirus is a dengue virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is selected from: DRGNGSGCGLNGS, DRGNGSGCGLFGK and DRGWGNGCGLNGS.
9. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, wherein the flavivirus is a Zika virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is DRNHTNGCGLFGK.
10. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses which is the product of expression of a recombinant nucleic acid sequence in a host cell capable of post-translational glycosylation.
11. An isolated recombinant analogue of a flavivirus E-protein of any one of the preceding clauses, which is the product of expression of a recombinant nucleic acid sequence in a host cell capable of glycosylation with an N-linked glycan.
12. An isolated recombinant analogue of a flavivirus E-protein of clause 10 or clause 11, wherein the host is a mammalian cell or insect cell.
13. An isolated recombinant analogue of a flavivirus E-protein of any one of clauses 10 to 12, wherein the host is a HEK cell line or a Tni cell line.
14. A diagnostic test or test kit comprising an isolated recombinant analogue of a flavivirus E-protein of any one of clauses 1 to 13 and a reagent capable of detecting an immunological (antigen-antibody) complex which contains said isolated analogue or binding molecule.
15. A diagnostic test or test kit according to clause 14, further comprising one or more control standards and/or a specimen diluent and/or washing buffer.
16. A diagnostic test or test kit according to clause 14 or 15, wherein said analogue and/or binding molecule is immobilized on a solid support.
17. A diagnostic test or test kit according to any one of clauses 14 to 16, wherein the solid support is a microplate well.
18. A diagnostic test or test kit according to any one of clauses 14 to 17, wherein an immunological complex which contains said isolated analogue or binding molecule is detected by ELISA.
19. A diagnostic test or test kit according to any one of clauses 14 to 18, wherein said immunological complex which contains said isolated analogue or binding molecule is detected by lateral flow.
20. A diagnostic test or test kit according to any one of clauses 14 to 19, wherein said test or kit comprises a test device comprising a lateral flow test strip comprising:
a sample pad for application of a liquid sample.
a conjugate pad comprising a detector conjugate for conjugation of anti-flavivirus antibody in the liquid sample,
a capture strip (e.g. nitrocellulose strip) comprising a capture means to capture the detector conjugate-anti-flavivirus antibody complex
and
an absorbent pad,
the pads and capture strip being arranged to permit capillary flow communication with each other.
21. A diagnostic test or test kit according to clause 20, wherein the sample pad comprises a red blood cell arresting agent, e.g. an anti-glycophorin antibody.
22. A diagnostic test or test kit according to clause 20 or claim 21 wherein the detector conjugate is a coloured particle (e.g. colloidal gold) conjugated to an anti-human antibody (e.g. anti-human IgG or anti-human IgM antibody) and the capture means for capture of the detector conjugate—anti-flavivirus antibody complex is an antigen comprising a recombinant analogue of a flavivirus E-protein of any one of claims 1 to 13.
23. A diagnostic test or test kit according to any one of clause 20 to 22, wherein the antigen is attached directly to the capture strip or indirectly via a tag system whereby an anti-tag reagent on the capture strip binds the tagged antigen.
24. A diagnostic test or test kit according to clause 20 or clause 21, wherein the detector conjugate is a coloured particle (e.g. colloidal gold) conjugated to an antigen comprising a recombinant analogue of a flavivirus E-protein of any one of clauses 1 to 13 and and the capture means for capture of the detector conjugate—anti-flavivirus antibody complex is an anti-human Ig antibody (e.g. anti-human IgG antibody or anti-human IgM antibody).
25. A diagnostic test or test kit according to clause 24, wherein the anti-human Ig antibody is attached directly to the capture strip or indirectly via tag system whereby an anti-tag reagent on the capture strip binds to the tagged antibody.
26. A diagnostic test or test kit according to clause 23 or 25, wherein the tag system is a His tag system, a FLAG tag system or a Streptavidin tag system.
27. A diagnostic test or test kit according to any one of clauses 20 to 26 wherein the liquid sample is a biological sample
28. A diagnostic test or test kit according to one of clauses 20 to claim 27, wherein the liquid sample is a biological sample selected from blood, plasma, serum, saliva and CSF.
29. A diagnostic test or test kit according to any one of clauses 20 to 28, wherein lateral flow test strip is housed within a casing, said casing having a window for visual inspection of the test result wherein during use the accumulation of colored particles produces a color indicative of the presence of an anti-flavivirus antibody in the liquid sample, said casing comprising a first port for application of the liquid sample or liquid sample and diluent, optionally comprising a second port (preferably more distal from the window than the first port) for application of the diluent.
30. A method for detection of a flavivirus antibody in a sample comprising use of a diagnostic test or test kit according to any one of clauses 14 to 29.
31. A diagnostic test, kit, test device or method substantially as described herein with reference to the description, drawings and/or clauses 1 to 30 above and to the claims.
The invention employs an isolated recombinant analogue of a flavivirus E-protein fusion loop comprising at least one glycosylation site for an N-linked glycan that is not present in a natural flavivirus E-protein fusion loop sequence, wherein the at least one glycosylation site is an N-linked glycosylation sequon (Asn-X-Ser/Thr) and the Asn (N) residue of the sequon may occupy any of positions 98-110 (SEQ ID NO: 1 DRGWGNGCGLFGK) of the natural flavivirus E-protein fusion loop amino acid sequence, wherein X is any amino acid residue except proline and Ser/Thr denotes a serine or threonine residue.
An isolated recombinant analogue of a flavivirus E-protein fusion loop may comprise two glycosylation sites that are not present in a natural flavivirus E-protein fusion loop sequence.
The invention employs an isolated recombinant analogue of a flavivirus E-protein comprising an analogue of a flavivirus E-protein fusion loop of the invention. In some embodiments the only modifications to the sequence of the isolated recombinant analogue of a flavivirus E-protein are the modifications in the fusion loop to introduce N-linked glycosylation sequon(s) (Asn-X-Ser/Thr), in other embodiments one or more further modifications may be introduced in flavivirus E-protein at residues outside the fusion loop.
An analogue of the having at least one glycan attached thereto is preferred. Preferably the at least one glycan is an N-linked glycan. Preferably the analogue is the product of expression of a recombinant nucleic acid sequence. At least one glycan may be present at one or more native glycosylation sites in the flavivirus E-protein outside the flavivirus E-protein fusion loop.
An analogue employed in the invention, may comprise an N-linked glycosylation sequon (Asn-X-Ser/Thr) such that an Asn (N) residue of the sequon occupies any of positions 98-101 and/or 106-110.
Preferably, X is any of the following 13 amino acid residues Gly, His, Asn, Gln, Tyr, Val, Ala, Met, Ile, Lys, Arg, Thr or Ser.
In preferred analogues for use in the invention, the flavivirus E-protein is a dengue virus E-protein and the Asn (N) residue of a sequon occupies position 101, 108 or both 101 and 108 of the amino-acid sequence of the analogue flavivirus E-protein fusion loop or the flavivirus E-protein is a Zika E-protein and the Asn (N) residue of a sequon occupies position 100 of the amino acid sequence of the analogue flavivirus E-protein fusion loop.
In a preferred analogues for use in the invention, the flavivirus is a dengue virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is selected from: SEQ ID NO: 2 DRGNGSGCGLNGS, SEQ ID NO: 3 DRGNGSGCGLFGK and SEQ ID NO: 4 DRGWGNGCGLNGS.
In another preferred analogue for use in the invention, the flavivirus is a Zika virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is SEQ ID NO: 5 DRNHTNGCGLFGK.
An isolated recombinant DNA or RNA sequence comprising a sequence encoding an analogue of a flavivirus E-protein fusion loop for use according to the invention.
An isolated recombinant DNA sequence may be a plasmid or a linear DNA-based vaccine. An isolated recombinant DNA sequence of the invention may encode an analogue of a flavivirus E-protein according to the invention under control of a mammalian promoter.
A host cell comprising a DNA or RNA sequence according to the invention may be an eukaryotic host cell comprising a DNA sequence according to the invention or a plasmid.
Preferably, a host cell is capable of expressing an analogue for use in the invention. Further preferably, a host cell is capable of expressing and glycosylating an analogue for use in the invention.
A method of making an analogue for use in the invention may comprise culturing a host cell according to the disclosure in conditions suitable for expression of the analogue and isolating the analogue.
Further provided is a composition comprising an analogue for use in the invention and a diluent.
A composition may comprise one or more flavivirus analogues of the invention selected from an analogue of DEN-1, an analogue of DEN-2, an analogue of DEN-3, an analogue of DEN-4 and an analogue of Zika.
A composition may comprise four dengue analogues of the invention representing each of the four dengue virus serotypes DEN-1 DEN-2 DEN-3 and DEN-4.
A composition may comprise a zika virus analogue of the invention.
A composition may comprise four dengue analogues of the invention representing each of the four dengue serotypes DEN-1 DEN-2 DEN-3 and DEN-4 and a zika virus analogue of the invention.
The disclosure also provides a binding molecule capable of binding specifically to an analogue described herein. The binding molecule may be an antibody or a fragment thereof, a domain antibody, a protein scaffold, or an aptamer, provided that it is capable of binding specifically to an analogue described herein.
In preferred embodiments the flavivirus infection is a dengue virus infection or a Zika virus infection.
The disclosure provides vaccine approaches that are designed purposefully to avoid the problem of antibody-dependent enhancement, the vaccine approaches employ an analogue, composition, binding molecule or diagnostic test of the invention in an in vitro method for diagnosis of flavivirus infection and/or to investigate exposure to flavivirus, to determine if a subject proposed for immunisation is naïve to Dengue and/or Zika infection and/or has been exposed to dengue and/or Zika infection, thereby to inform the decision to immunise against dengue and/or Zika if the subject is naïve to dengue and/or Zika infection, or not to immunise if prior exposure to dengue and/or Zika is detected.
The invention provides an analogue, composition or binding molecule of the invention for use as a diagnostic.
The invention provides a diagnostic kit comprising an analogue, composition or binding molecule of the invention and a reagent capable of detecting an immunological (antigen-antibody) complex which contains said isolated analogue or binding molecule.
A diagnostic test kit in accordance with the invention may further comprise one or more control standards and/or a specimen diluent and/or washing buffer.
In a diagnostic test kit of the invention, the analogue and/or binding molecule specific thereto of the invention may be immobilized on a solid support. The solid support may be a microplate well. In a diagnostic test kit according to the invention, an immunological complex which contains said isolated analogue or binding molecule may be detected by ELISA or by lateral flow.
The invention provides diagnostic approaches that can differentiate between closely-related Flaviviruses, to assess if an individual is seronegative and thus has not been exposed to dengue or Zika, or if an individual is seropositive and has been exposed to Zika and/or dengue and for those who are seropositive, to distinguish to which of Zika and/or Dengue the individual has been exposed, in some aspects it will be distinguished to which of the four dengue serotypes the individual has been exposed. The invention provides diagnostic approaches that can be used to select subjects for immunization, or assess seroconversion to determine if immunization has raised a protective immune response against dengue or Zika. The invention provides diagnostic approaches that enable interrogation of the immune response to distinguish antibodies against the dengue virus serotypes and against Zika virus.
As described herein, we have developed and exemplified (in ELISA and lateral flow studies) the concept that diagnostics for Zika and dengue (and other flaviviruses) can be improved by creating antigens of enhanced type-specificity and serotype-specificity by cloaking the immunodominant fusion loop of the envelope protein with a glycan (these hyperglycosylated exodomain antigens are termed ‘HX’, in order to denote the presence of one or more additional glycans in the fusion loop). The fusion loop of dengue virus represents a 14-residue sequence element, comprising just a small percentage of the envelope protein surface, yet is the target of 55% of antibodies following a primary infection (Dejnirattisai et al., 2014) and >90% of antibodies following a secondary infection with dengue (Beltramello et al., 2010; Lai et al., 2008) (detailed figures for Zika are not yet available). Moreover, as explained above, the fusion loop sequence is 100% conserved across all four dengue serotypes, Zika, yellow fever, West Nile and Japanese encephalitis viruses, such that antibodies against the fusion loop are highly cross-reactive, not just between dengue and Zika but across these other more-distantly related viruses also. We have demonstrated herein that our approach of cloaking the fusion loop preserves the antigenic structure of Zika and dengue viral proteins, retaining conformational and neutralising epitopes while abolishing fusion loop reactivity (which, stereotypically, is dominated by the surface-exposed hydrophobic residues of the fusion loop). The glycan(s) that we have introduced into the fusion loop of the dengue and Zika envelope proteins transform the natural topography of this structure in a more profound way than amino acid replacements alone are capable of, supplanting a strongly hydrophobic surface patch (a contiguous surface formed of the hydrophobic side chains) with large branched hydrophilic glycan structures orthogonal to the path of the polypeptide chain. As shown herein, these modifications rigorously prevent recognition by fusion loop antibodies. In ELISA tests we found that the cloaked antigens have equal sensitivity (to wild-type equivalent antigens) for the detection of antibodies against multiple non-fusion-loop epitopes. Moreover we demonstrated that wild-type post-Zika macaque and tamarin sera strongly recognize the (uncloaked) fusion loop of wild-type dengue-2 and dengue-4 exodomain antigens, demonstrating that the fusion loop is likewise an immunodominant element of the Zika virus, in the course of natural infection (at least in the case of these non-human primate species), and a major cause of off-target recognition by antibodies in convalescent sera. However, there was no off-target reactivity of convalescent Zika primate sera with HX versions of the dengue envelope antigens. Furthermore, as expected, the HX Zika antigen was strongly recognized by primate Zika convalescent sera. These observations demonstrated the superior diagnostic sensitivity and specificity of HX antigens in ELISA tests, compared to wild-type antigen. We describe the translation of the advantages of these novel HX proteins to a lateral flow ‘pregnancy-test-like’ format for point-of-care diagnostic use.
The invention is be described with reference to various embodiments of different aspects of the invention. It is appreciated that certain features of the invention, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in one or more embodiments or in a single embodiment. Conversely, various features of the invention, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination. All combinations of the embodiments are specifically embraced by the present invention and are disclosed herein just as if each and every combination was individually and explicitly disclosed. In addition, all sub-combinations are also specifically embraced by the present invention and are disclosed herein just as if each and every such sub-combination was individually and explicitly disclosed herein.
The invention uses modified Flavivirus nucleic acid and protein sequences in which the natural (native, wild-type) E-protein fusion loop epitope, known to be associated with generation of flavivirus cross-reactive, infection-enhancing antibodies has been modified to comprise one or more (e.g., 2) glycosylation sites for glycosylation of the protein with an N-linked glycan that is not normally present on the native fusion loop epitope. Such modification alters the fusion loop amino acid sequence and the presence of a glycan further disguises the epitope. Thus the modified Flavivirus nucleic acid and protein sequences of are designed to generate a protective response without concomitant generation of flavivirus cross-reactive infection-enhancing antibodies, thereby intending to avoid the problems of antibody-dependent enhancement observed with existing vaccine approaches. The modified Flavivirus nucleic acid and protein sequences are also designed for diagnostic use, either as antigens for detection of a specific Flavivirus or to generate binding molecules such as antibodies for detection of a specific Flavivirus.
By antibody we include the meaning of a substantially intact antibody molecule, as well as a chimeric antibody, humanised antibody (wherein at least one amino acid is mutated relative to a non-human antibody, for example a naturally occurring non-human antibody or antibody assembled from non-human antibody sequences), single chain antibody, bi-specific antibody, antibody heavy chain, antibody light chain, homo-dimer or heterodimer of antibody heavy and/or light chains, and antigen binding portions and derivatives of the same.
A binding molecule of the invention is preferably an antibody or antigen binding portion thereof. An antibody may be a full antibody with Fc, or antigen binding portion thereof. The antigen binding portion may be a Fv fragment; a Fab-like fragment (e.g. a Fab fragment, a Fab′ fragment, a F(ab)2 fragment, Fv or scFv fragments); or a domain antibody. The antigen binding portion may be derived from the linear amino acid sequence present in an intact antibody, or may comprise a set of non-consecutive amino acids, optionally interspersed with other amino acids, for example may comprise particular amino acids that are required for contact with an epitope, but may for example not comprise the amino acids required for the framework of a native antibody, which, in some cases, may be replaced by a heterologous scaffold protein, for example. An antibody according to the present invention is obtainable by a method comprising a step of immunizing a mammal, such as a human, a monkey, a rabbit or a mouse; and/or by an in vitro method, for example comprising a phage display selection step, as will be well known to those skilled in the art.
The term antibody also includes all classes of antibodies, including IgG, IgA, IgM, IdD and IgE. The term antibody also includes variants, fusions and derivatives of any defined antibodies and antigen binding portions thereof.
By neutralise we mean reduce the ability of the virus to infect previously uninfected cells. The person skilled in the art will be well aware of suitable techniques to monitor viral neutralising ability.
Methods for manipulation of nucleic acid sequences to introduce sequence changes as described herein are well known in the art.
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
DRGWGNGCGLFGK
The fusion loop DRGWGNGCGLFGK (defined as residues 98-110, SEQ ID NO: 1) in the wild type sequences (rows 1 to 9) is shown in bold. The residues changed to make the N-linked glycosylation sequons in the modified analogue HX sequences are shown in bold in rows 10-20 The constructs pCRO21-24, 26, and 28 expressed well and were selected for further investigation. In the case of dengue E-proteins, 4 residues were changed to make two glycosylation sites (pCRO21-24). In the case of Zika E-protein, 3 residues were changed to make one glycosylation site (pCRO28).
The constructs pCRO25, 29, 30 and 31 did not express well in the expression system chosen, thus in some contexts the recombinant analogue sequences of the invention do not comprise the following sequences:
In an analogue of the invention, the N-linked glycosylation sequon (Asn-X-Ser/Thr) may be present such that an Asn (N) residue of the sequon occupies any of positions 98-101 and/or 106-110. That is, the N residue may occupy position a position selected from 98, 99, 100, and 101 and/or a position selected from 106, 107, 108, 109 and 110.
Preferably, in an analogue of the invention, X is any of the following 13 amino acid residues Gly, His, Asn, Gln, Tyr, Val, Ala, Met, Ile, Lys, Arg, Thr or Ser, with Gly or His being particularly preferred. In specific embodiments of the invention described herein for dengue viruses it is preferred that X is Gly and for Zika is preferred that X is His.
In preferred analogues of the invention, the flavivirus E-protein is a dengue virus E-protein and the Asn (N) residue of a sequon occupies position 101, 108 or both 101 and 108 of the amino-acid sequence of the analogue flavivirus E-protein fusion loop or the flavivirus E-protein is a Zika E-protein and the Asn (N) residue of a sequon occupies position 100 of the amino acid sequence of the analogue flavivirus E-protein fusion loop.
In a preferred analogue of the invention, the flavivirus is a dengue virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is selected from: DRGNGSGCGLNGS (SEQ ID NO: 2), DRGNGSGCGLFGK (SEQ ID NO: 3) and DRGWGNGCGLNGS (SEQ ID NO: 4).
In another preferred analogue of the invention, the flavivirus is a Zika virus and the amino acid sequence of the analogue flavivirus E-protein fusion loop 98-110 is DRNHTNGCGLFGK (SEQ ID NO: 5).
The nucleic acid sequence encoding recombinant analogue E-protein fusion loop protein or encoding recombinant analogue E-protein comprising such fusion loop protein can be generally be expressed following the functional and operable insertion of the DNA sequence into an expression vector containing control sequences and secretory signal sequences.
A suitable promoter for expression of nucleic acid sequences of the invention is CMV for expression in mammalian cells.
Host cells that may be employed in accordance with the invention include the mammalian HEK and CHO cell lines, insect cells such as Tni and Drosophila S2; yeasts and non-mycelial fungi cells. The host may be genetically engineered to produce therapeutic glycoproteins with human-like N-linked glycans.
The practice of the present invention will employ, unless otherwise indicated, conventional techniques of molecular biology, microbiology, recombinant DNA, and immunology, which are within the skill of the art. Such techniques are explained fully in the literature. See e.g., Sambrook, Fritsch, and Maniatis, Molecular Cloning: A Laboratory Manual, Second Edition (1989), Oligonucleotide Synthesis (M. J. Gait Ed., 1984), Animal Cell Culture (R. I. Freshhey, Ed., 1987), the series Methods in Enzymology (Academic Press, Inc.); Gene Transfer Vectors for Mammalian Cells (J. M. Miller and M. P. Calos eds. 1987), Handbook of Experimental Immunology, (D. M. Weir and C. C. Blackwell, Eds.), Current Protocols in Molecular Biology (F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Siedman, J. A. Smith, and K. Struhl, eds., 1987), and Current Protocols in Immunology (J. E. Coligan, A. M. Kruisbeek, D. H. Margulies, E. M. Shevach and W. Strober, eds., 1991). All patents, patent applications, and publications mentioned herein, both supra and infra, are hereby incorporated herein by reference.
Standard three and one-letter terminology is used for amino acid residues.
As used herein, the term “recombinant” refers to the use of genetic engineering methods (cloning, amplification) to produce an analogue, or a binding molecule such as an antibody or an antibody fragment of the present invention.
The principal problem of dengue vaccine development, wherein the use of vaccines runs the risk (in a finite number of cases) of giving rise to ‘antibody dependent enhancement’ of dengue infection, making the illness worse rather than preventing it. Enhancement is a feature of natural infection (where antibodies sent to neutralize the virus are subverted to gain access to human myeloid cells), usually upon encounter with a second ‘serotype’ of virus, resulting in more severe symptoms (Halstead, Rojanasuphot, & Sangkawibha, 1983). Vaccination, while for the most part conferring protection, is also liable on some occasions to predispose a recipient to severe dengue, including dengue haemorrhagic fever (DHF), upon first exposure to a wild dengue virus: i.e., ‘iatrogenic’ cases of severe dengue or DHF, which would not have occurred but for the vaccine. Furthermore, existing vaccine approaches also have the potential to create a population of vaccinated individuals who develop severe iatrogenic dengue, at some interval after the vaccine (or vaccine course) has been administered (e.g. a decade). This is because, as immunity to dengue wanes, protective antibodies reach a concentration where they ‘enhance’ rather than prevent infection. Also, the rate of decay of ‘immunological memory’ (where the immune system recalls encounter with a wild virus or vaccine dose) is not synchronous for the four serotypes of the vaccine, such that immunity to each serotype (at the antibody and memory level) of dengue is lost at different times, successively increasing the risk of severe disease. This gradual failure of immune memory likewise creates a new population of individuals who are now predisposed to severe dengue (when bitten by an infected mosquito), instead of protected, as a result of previous vaccination. While current vaccines against dengue (licensed and in development) may meanwhile prove to be of substantial ‘net’ benefit to public health, improved safety is still desirable in order to avoid cases of vaccine-induced dengue (i.e., iatrogenically-caused severe dengue). The likely role of natural dengue infection in paving the way for pandemic Zika infection has been elaborated recently by Philip K Russell of the Sabin Vaccine Institute (Russell, 2016). While no systematic investigation has been conducted that would determine the risk of dengue vaccination predisposing to Zika virus infection or of dengue vaccination giving rise to Zika infections of enhanced severity, it is a logical extension of Russell's observations to expect such cases. Likewise although dengue-vaccine-induced predisposition to severe dengue has not yet been reported or investigated ‘as such’, in a recent three-year follow-up study of the Sanofi-Pasteur vaccine there was an increased rate of hospitalisation in children less than nine years of age (Hadinegoro et al., 2015) which could be explained by vaccine-induced enhancement of susceptibility to severe dengue. These new epidemiological developments, and laboratory data (below) indicate that there is a significant risk that vaccines (unless designed to avoid enhancement) will cause, in some instances, enhancement of disease: i.e. dengue vaccination will result in cases of severe dengue that would not otherwise have happened. It is also possible that dengue vaccines could facilitate the spread of Zika virus infection if used on a population-wide basis. The legitimacy of this concern is supported additionally by in vitro experimental data which demonstrates that dengue virus antibodies enhance the infection of human myeloid cells by Zika virus (Paul et al., 2016). Furthermore, it follows that a stand-alone Zika vaccine could give rise to similar antibodies that would (conversely) enhance dengue infection giving rise to cases of severe iatrogenic dengue, by generating anti-Zika antibodies that cross-react with dengue virus, and that facilitate dengue infection. For the purposes of this application, while not wishing to be bound by any particular hypothesis, Zika virus is accorded the status of a ‘fifth dengue serotype’. This is because dengue infection (and dengue vaccines) have the potential to facilitate the spread of Zika by generating infection-enhancing antibodies which also react with Zika virus facilitating its infection of bodily cells.
Since the advent of Zika as a pandemic phenomenon, its rapid global spread apparently facilitated by dengue-infection (Russell, 2016), the problem of vaccination (i.e. how to make a vaccine that does not, in some cases, worsen disease) has become more complicated. A new vaccine design is required in order to avoid homologous enhancement (whereby a dengue vaccine would facilitate, in some cases, dengue infection) and cross-enhancement (whereby a dengue vaccine would facilitate, in some cases, Zika infection); and moreover, whereby a Zika vaccine would facilitate, in some cases, dengue infection. Conventional approaches to the antibody enhancement problem, which involve such stratagems as combining all four serotypes of dengue in a single vaccine (Sanofi-Pasteur) or, for example, a subunit approach using N-terminal regions of the E-proteins of dengue (Merck) have recognized the antibody enhancement problem but have not provided a comprehensive solution appropriate to the Zika-pandemic situation. The most advanced dengue vaccine (the licensed Sanofi-Pasteur live attenuated tetravalent dengue vaccine), fails to deal with Zika, and from the epidemiological and in vitro observations above may be capable of promoting cases of Zika virus infection by cross-enhancement (even while having a net benefit community-wide by dint of herd immunity).
It is important to recognize that the distinction between enhancing epitopes and protective epitopes of flaviviruses is not ‘binary’ in character. Generally speaking, almost all anti-dengue-E antibodies (for example) have the potential to be both neutralising and infection-enhancing, the latter property emerging at lower antibody concentrations (Dejnirattisai et al., 2014), e.g. as immunity to a vaccine or an exposure wanes. Moreover, Dejnirattisai et. al. also found that antibodies against the fusion loop of the dengue E-protein (which comprise about half of all antibodies generated convalescently) are markedly worse than antibodies against other sites on the E-protein in terms of their propensity for antibody-dependent enhancement of infection.
The present disclosure provides a diagnostic test that may help to mitigate the issues of antibody-dependent enhancement and cross-enhancement, by detecting and distinguishing prior Zika and dengue virus infections to allow the intelligent and safe development and deployment of Zika and dengue vaccines.
Tests of the invention may be ‘serologically’ based (measuring antibodies) so as to determine whether a person had previously had Zika or dengue infection, because the infection and vaccination history of a subject is an important determinant of how they will respond to new vaccines recently licensed (i.e. Dengvaxia) and in development (Zika vaccines), and may determine a subject's susceptibility to adverse responses to vaccination against dengue and Zika, such as predisposition to dengue haemorrhagic fever and dengue shock syndrome.
Diagnostic tests of the invention can be used to identify prior Zika or dengue infection by detecting antibodies in a fingerprick blood sample.
Diagnostic tests of the invention may be provided as a lateral flow device for detection of antibodies against Zika and dengue viruses, or a pair of lateral flow devices for measuring antibodies against Zika and dengue viruses, respectively. The tests of the invention will help define which subjects are best-suited to receive Zika vaccination, maximising the safety and effectiveness of Zika vaccine deployment. The tests also detect virus-neutralising antibodies and will also help define whether a subject has responded adequately to the vaccine (i.e. reached a ‘to be determined’ protective level of Zika-specific antibodies) or whether they may require a further dose (a factor that may vary in differing endemic territories, depending on prior exposure to related viruses). Also, since the near-withdrawal of Dengvaxia (which is no longer permitted for use in persons not previously infected with dengue virus), the new test, once licensed, will allow safer deployment of Dengvaxia (and future dengue vaccines), markedly reducing the risk of ‘priming’ for haemorrhagic fever by (inadvertent) vaccination of dengue-naïve subjects.
The diagnostic of the invention can be used to distinguish Zika and dengue viruses to allow the intelligent and safe development and deployment of Zika and dengue vaccines. The test can be ‘serologically’ based (measuring antibodies) so as to determine whether a person had previously had Zika or dengue infection, because the infection and vaccination history of a subject is an important determinant of how they will respond to new vaccines recently licensed (i.e. Dengvaxia) and in development (Zika vaccines), and may determine a subject's susceptibility to adverse responses to vaccination against dengue and Zika, such as predisposition to dengue haemorrhagic fever and dengue shock syndrome. While other tests are available that are quite specfic (eg. PCR), the disadvantage of these tests is that they only work during the very brief active phase of the infection (about seven days) which is easily missed, and PCR tests are not adequate to inform risk of adverse reactions to vaccination because dengue and Zika infections are frequently asymptomatic, and would not (in the ordinary course of events) be sampled for PCR.
There is an urgency to accurately monitor, map and contain outbreaks of Zika, which, while mostly asymptomatic, is easily confused, symptom-wise, and in antibody-based tests, with various other infections. The diagnostic test of the invention will enable these new vaccines to be used most effectively. Thus, once vaccines become available against Zika, it will be important to deploy precious vaccine resources appropriately in the field in resource-limited countries, making sure the vaccine is deployed to ‘current’ outbreak-areas as a priority, as the geographic prevalence is dynamic and changes over time. This will save health-care costs on unnecessary use of vaccine. In this way the diagnostic tests of the invention can play an important role in monitoring Zika activity and distinguishing it from dengue and other clinically similar infections.
Also, the diagnostic tests of the invention can be used to monitor the effectiveness and safety of development/deployment of novel vaccines and to understand the risks of vaccine development against this ‘new’ virus (i.e., Zika virus), which is only one of a group of related co-endemic viruses, including dengue, West Nile, Japanese encephalitis and yellow fever viruses (depending on the territory).
It is becoming increasingly apparent that antibodies resulting from prior infections (or vaccinations) with related viruses, especially dengue, already present in the blood of patients experiencing Zika virus infection, may influence the course of Zika disease and its complications. Zika infection has been implicated as a causative factor, via foetal infection, of microcephaly in neonates, as well as Guillain-Barré syndrome in adults). Likewise, antibodies generated during Zika infection or as a consequence of Zika vaccination may have analogous pathological consequences to those elicited by dengue infection and vaccination. For example, such antibodies may determine whether naturally-encountered Zika virus crosses the placenta and damages the foetus, and antibodies generated by Zika vaccination may influence the occurrence of haemorrhagic fever upon first or second encounter with dengue.
Increasingly, in the case of dengue, antibodies are seen as essential co-factors in severe disease causation, as important as the viruses themselves. In the case of Zika this may also transpire to be the case. That Zika infection (or vaccination) may prime for dengue haemorrhagic fever is already strongly indicated by in vitro experiments with human cells and by animal in vivo experiments. The extent to which Zika antibodies may be a problem in man will take years of epidemiological research to unravel, and this will require tests of improved sensitivity and specificity—such as the tests of the invention.
Current diagnostic practice is not adequate to properly inform the intelligent deployment of vaccines because it is based in central diagnostic labs that require the transport of blood samples from diverse, sometimes remote, areas of endemic countries, which is expensive and inefficient, and which may not be possible in many instances due to lack of the necessary refrigerated transport infrastructure. Also, central and point-of-care serological tests developed for dengue in the pre-Zika era (ie. the only currently licensed tests) are confounded by Zika cross-reactivity with error-rates of 50% or more, making these tests of limited use in distinguishing Zika from dengue.
We have therefore developed a ‘point-of-care’ diagnostic test of the invention, i.e., a test that can be run without the need for clean water, electricity or equipment, that can be operated in the home or in a vaccination clinic, that can distinguish prior Zika or dengue infection with improved reliability over existing point-of-care tests (and improved relative to existing central laboratory-based tests). The test may comprise a test for Zika and or a test for and dengue, these may be provided individually as test devices or as part of a single diagnostic test device. The device(s) may be lateral flow device(s). In a preferred embodiment the test comprises a pair of lateral flow devices—one each for Zika and dengue. The diagnostic tests of the invention are enabled by the design of antigens engineered to render the immunodominant site of the viral envelope proteins ‘invisible’ to antibodies by the strategic planting of glycans in the fusion loop of the E protein, which is a small, highly conserved (100% conserved) site, recognized by the majority of antibodies that are generated in the course of dengue and Zika infections.
The diagnostic tests are based on recombinant analogues of Zika and dengue envelope (E) protein fusion loops into which one or more glycosylation sites have been introduced. The present invention uses E-proteins with at least one an additional glycan planted in the fusion loop, by virtue of engineering an additional, novel, glycosylation site into the nucleotide and amino acid sequence of recombinantly expressed E-proteins. The ‘cloaking’ effect of the glycan prevents cross-reactive (ADE-enhancing) antibodies binding to the fusion loop site, while leaving other sites that bind to neutralising antibodies available for binding to enable specific detection of neutralising antibodies (and thus evidence of prior virus exposure).
The E-protein of Zika virus is highly homologous in terms of its amino acid sequence and three-dimensional structure, to that of the dengue virus E-proteins. The recent cryo-EM 3.8 Angstrom structure of the Zika virion E-protein clearly identifies (by analogy) the Zika E-protein fusion loop location (Kostyuchenko et al., 2016; Sirohi et al., 2016). Indeed Sirohi et. al. catalogue the remarkable degree of homology among diverse flaviviruses with respect to the fusion loop sequence “DRGWGNGCGLFGK” (residues 98-110), which is perfectly preserved among diverse virus isolates of Zika, the four dengue serotypes, West-Nile, Japanese encephalitis and yellow fever viruses (see supplementary figure S2 of Sirohi). There are notable differences between dengue and Zika E-proteins, such as a five amino acid insert in the Zika E-protein, and the fact that Zika has a single N-linked glycan rather than two per monomer
Methods for introducing additional glycosylation sites into proteins by site directed mutagenesis are well known in the art. In particular the creation of Aranesp (darbepoetin alfa), a modified form of the natural hormone erythropoietin, is a good example (Elliott (“EP0640619A1,” 2010), (Elliott et al., 2003). It is important in making suitable genetic constructs to ensure that the leader sequence of the protein is incorporated into recombinant plasmid or other vector DNA sequences, in order to direct the nascent polypeptide chain into the endoplasmic reticulum of the host cell, allowing glycosylation and to facilitate protein folding. Various eukaryotic cell systems are suitable for recombinant production—such as Chinese hamster ovary cells (CHO), as well as yeast (e.g., Pichia pastoris) and other vector systems such as baculovirus (which has the added advantage of equipping the viral protein immunogen with an insect glycan, as per the inoculum form of the flavivirus). However, prokaryotic systems such as those based on E. coli are not generally suitable, because they do not have the cellular apparatus required to effect glycosylation of proteins. In the case of Aranesp, the molecule has two additional N-linked glycosylation sites, strategically placed to avoid hindrance of interaction of the glycoengineered molecule with the erythropoietin receptor. The purpose of glycoengineering the earlier erythropoietin-based product in this way was to improve the longevity of the molecule in circulation by increasing its size giving rise to a product that can be administered once instead of thrice weekly (Elliott et al., 2003).
In studies (by others) with human monoclonal antibodies isolated from patients previously exposed to dengue infection, 100% of (monoclonal) antibodies isolated in the first month of a Zika infection were found to be cross-reactive with dengue. Meanwhile, the need for an IgG-based assay for Zika was asserted by the observation that previously dengue infected subjects (90-95% of the population in Rio, for example), may bypass altogether a Zika-IgM response (the usual indicator of recent infection) launching instead an anamnestic ‘boost’ response of cross-reactive IgG antibodies, that gradually mature to include Zika-specific antibodies. Because Zika is asymptomatic in 80% of cases, measurement of IgG antibodies is the only way to obtain an accurate estimate of the burden of infection in society (IgM antibodies, when produced, being too transient for the purpose). The Zika IgG test of the invention will be instrumental in the management of patients presenting late, after an asymptomatic infection, with neurological complications or with congenital malformations to see if these phenomena indicate the effect of a prior Zika virus infection.
The invention will now be described with reference to the accompanying drawing in which:
The x-axis shows the number of days after immunisation and the y-axis shows the IgG antibody titre. Three doses were given on days 0, 14 and 21. Dosages are indicated in Table 9. Antibody responses were measured in individual mice against all five antigens as wild-type VLPs on the ELISA solid phase as indicted: top row left Den 1 VLP antigen, top row right Den 2 VLP antigen, middle row left Den 3 VLP antigen, middle row right Den 4 VLP antigen, bottom row left Zika VLP antigen. Immunogens (as distinct from antigens uses for assay above) were Penta-DNA (a combination of each of the Den1-4 and Zika DNAs of the invention) shown as an open circle, Penta-Prot (a combination of each of the Den1-4 and Zika proteins of the invention) is shown as an filled square, Monovalent Zika is shown as a filled triangle, Penta VLP (a combination of each of the Den1-4 and Zika VLPs of the invention) is shown as a filled inverted triangle. PBS control is shown as an open inverted triangle.
In order to further characterize the hyperglycosylated antigens of the present disclosure, comparing them to wild-type equivalent antigens, an ELISA assay was established to measure antibody binding to diverse wild-type and recombinant exodomains (as distinct from the VLP antigens of
Upper panel shows ELISA reactivity of antibodies in a dengue convalescent serum with immobilized Zika and dengue wild-type (WT) and hyperglycosylated (HX) exodomain proteins oriented on the solid phase by capture with a rabbit anti-His-tag monoclonal antibody, in the presence (grey bars, right of each pair) and absence (black bars, left of each pair) of competing mouse monoconal flavivirus fusion loop antibody 4G2 (an anti-dengue-serotype-2 cross-reactive monoclonal antibody) at a concentration of 10 ug/ml during serum incubation. Human sera were tested at a constant concentration of 1/1000.
Lower panel shows ELISA reactivity of antibodies in a Zika convalescent serum with immobilized Zika and Dengue wild-type (WT) and hyperglycosylated (HX) exodomain proteins in the presence (grey bars) and absence (black bars) of competing mouse monoclonal flavivirus fusion loop antibody 4G2. Conditions and labelling are the same as for the upper panel. Error bars are standard error of duplicate determinations.
(B) With a high threshold (set at greater than 3=positive), Zika LF of the invention markedly outperforms commercial (Euroimmun/Perkin Elmer) and bespoke (DABA) Zika ELISA assays based on measuring the NS1 protein (a test for acute Zika infection), the NS1 ELISA tests score half of this BOB-negative sample group as positive for Zika (an implausible result given the sample collection date—before significant Zika circulation had occurred) whereas the Excivion Zika LF scores 2/50=4% as positive−a more plausible result (Zika was present in Brazil in 2014 but the precise exent of its circulation is unknown);
(C) dengue LF test of the invention (setting a threshold of greater than 3=positive, as in B) scores 92% of Rio 2014 endemic sera as dengue+ve, which tallies with the 90-95% accepted value based on several published studies of dengue seropositivity in Brazil at this time. The dengue LF of the invention tallies well 92% compared to the 90-95% dengue seropositivity in this 2014 sample group from Brazil, demonstrating excellent sensitivity of the dengue-LF (confirming results from the Pune tests, which demonstrated superior sensitivity of IgG detection over the Standard Diagnostics Duo LF test, p<0.0000001).
(D) Frequency-distribution of LF scores in the Zika LF and dengue LF tests of the invention in the Rio 2014 sample set (before significant Zika circulation had occurred); one subject had a zero score in the dengue LF test and exemplifies the utility of the test in identifying dengue-negative subjects for companion-diagnostic use of the test with dengue vaccines (e.g. Dengvaxia), i.e., in order to avoid priming for dengue haemorrhagic fever this subject would be spared vaccination, the two subjects who were LF-positive for Zika in this sample set (scoring ‘4’) may represent genuine Zika cases and the ability of the Zika LF of the invention to be used as a ‘sentinel test’ providing early warning of a Zika outbreak.
Conversely, ‘B’ illustrates a vaccine immunogen designed in accordance with the invention. The novel immunogen, containing an E-protein wherein the fusion loop sequence has been modified and has been designed to be substituted with a glycan with the aim to generate neutralising antibodies against the E-proteins of the vaccine without generating infection-enhancing antibodies. ‘D’ represents occasional failure of the vaccine of the invention to elicit a protective level of antibody response in some subjects (e.g., the immunosuppressed), however, unlike other vaccine designs known in the art, the vaccine of the invention is designed to not render immunosuppressed subjects susceptible to enhanced infection with dengue or Zika viruses. Immunogens and vaccines of the present design are thereby designed to be safer on an individual subject basis and moreover to lack the potential to facilitate the epidemic spread of Zika by creating a population of subjects that have Zika-infection-enhancing antibodies, in the absence of neutralising antibodies. (WT=wild type).
Example 2 (
Plasmid inserts encoding various novel recombinant forms of the natural wild type (WT) exodomain sequences representative of the four dengue serotypes and of Zika and containing an E. coli origin of replication and a cytomegalovirus (CMV) promoter, as well as a hexahistidine C-terminal tag, were made by de novo gene synthesis (Thermofisher, GeneArt). Where two glycosylation sequons were inserted in the DNA sequence, the sequence was changed ‘manually’ to avoid the creation of direct DNA sequence repeats that might otherwise allow undesirable homologous recombination events.
Plasmid expression vectors pCRO21 (SEQ ID NO: 13), pCRO22 (SEQ ID NO: 14), pCRO23 (SEQ ID NO: 15), pCRO24 (SEQ ID NO: 16) and pCRO28 (SEQ ID NO: 17), coding for the mutated exodomain of the Envelope proteins of DENV1, DENV2, DENV3, DENV4 and ZIKV, respectively, were ultimately selected and produced by The Native Antigen Company, Oxford, as follows: expression cassettes were synthesized de novo to contain a 5′ NotI site followed by a consensus Kozak sequence followed by the coding sequence for the first 17 amino acids of the influenza-A virus haemagglutinin protein acting as secretion signal. The Envelope protein coding sequences used, (numbering relative to the polyprotein), were 280-675 (NCBI ACA48859.1), 281-676 (NCBI ADK37484.1), 281-673 (NCBI AIH13925.1), 280-675 (NCBI ANK35835.1) and 291-696 (NCBI ARB07957.1), respectively. [Elsewhere, for ease of reference, numbering is expressed according to residue number in the E-protein, with W at 101 of the fusion loop as a reference point]. Each construct contained coding sequences for a glycine-serine linker 7 to 8 amino acids in length followed by a 6×His-tag and a stop codon. The stop codon is followed by a NheI site in each expression cassette. The mammalian expression vector pSF-CMV (Oxford Genetics, Oxford) was digested with NotI and NheI, and the 4.2 kb fragment was ligated to the 1.3 kb NotI and NheI fragments of the expression cassette harbouring maintenance vectors (pUC57). In each case, one or two additional sequons of the general formula (NXS/T) was introduced into the fusion loop of the E-protein exodomain, capable (theoretically) of encoding a functional N-linked glycosylation site. The wild-type dengue proteins naturally already have two glycosylation sites, and Zika one. None of the natural glycans are found in the fusion loop.
For small-scale preparation 15 ml aliquots of HEK293FT cells at 3e6/ml were individually transfected with pCRO21, pCRO22, pCRO23, pCRO24 or pCRO25 (SEQ ID NO: 18), 4 control transfections were performed using pSF233, pSF236, pSF237, pSF238 or pSF239. After a day, 15 ml of rescue medium was added to each transfection. At day 3 after transfection each of the 10 transfections was treated the same way as follows: 30 ml of suspension was spun at 4,000 g for 7 minutes. The resulting supernatant was filtered using a 0.22 um disc filter. The pellet was resuspended in 1 ml of PBS. The filtered supernatant was then concentrated using a Vivaspin20 (30,000 Da cutoff) as per manufacturer's instructions. Concentrate volumes ranged from 0.6 ml to 1.2 ml. All concentrates were brought up to 1.2 ml with PBS. The concentrated supernatants were subjected to Talon purification as per manufacturer's instructions using Talon HiTrap Spin (GE). Buffers for Talon capture were: Equilibration Buffer: 50 mM phosphate pH7.8, 300 mM NaCl; Wash Buffer: 50 mM phosphate pH78, 300 mM NaCl, 5 mM imidazole; Elution Buffer: 50 mM phosphate pH7.8, 300 mM NaCl, 150 mM imidazole.
Characterisation of the resulting proteins by coomassie-blue staining (
For scale-up production, the novel hyperglycosylated proteins were expressed recombinantly in human embryonic kidney cells (HEK 293) by transient transfection with linear polyethyleneimine (PEI), and purified by metal chelate affinity chromatography with a cobalt chelate (TALON®, Clontech/GE), as described as follows for the dengue-1 hyperglycosylated construct based on pCRO21. 20×1 L of HEK293 cells were transfected with DEN V1_Eexo_2 xglyco expression vector pCRO21. 3 days post transfection, the supernatant was harvested by centrifugation, and the cleared supernatant was 0.2 um filtered and concentrated to ˜200 ml by tangential flow filtration (TFF). Immobilised metal affinity chromatography (IMAC) was performed on the TFF retentate using 5 ml HiTRAP Talon pre-packed column (GE) according to manufacturer's instructions using 20 mM sodium phosphate pH7.8 based buffer systems. DEN V1_Eexo_2 xglyco protein containing fractions were pooled and dialysed against 20 mM TRIS-HCl pH7.8 10 mM NaCl. Ion exchange chromatography was performed using a pre-packed 5 ml HiTrap Q HP column according to manufacturer's instructions. DENV1_Eexo_2 xglyco were pooled and dialysed against DPBS pH7.4. The dialysed solution was 0.22 um filtered and vialled under sterile conditions. BCA assay and SDS-PAGE were performed according to manufacturer's instructions (Bio-Rad).
Note that three of the hyperglycosylated constructs express at levels much higher than wild type (these are the hyperglycosylated dengue serotypes 2, 3 and 4 corresponding to plasmids pCRO22, pCRO23 and pCRO24). Zika plasmid, pCRO25 did not give rise to detectable secreted protein (
Therefore a further round of constructs was made (see
The hyperglycosylated forms chosen were pCRO21, pCRO22, pCRO23, pCRO24 (for dengue serotypes 1-4 respectively) and pCRO28 for Zika. Hyperglycosylated exodomains D1, D2, D3, D4 and Zika correspond to plasmids pCRO21, pCRO22, pCRO23, pCRO24 and pCRO28, respectively (SEQ ID NO: 24, 25, 26, 27 and 28 respectively). Molecular weight increments due to glycosylation are apparent, higher for the +2 dengue constructs than for the Zika +1 construct.
In all, eleven plasmid constructs were made and tested for protein expression and five were selected for further investigation, based on equivalent or (in most cases) superior levels of expression compared to wild type (pCRO21, pCRO22, pCRO23, pCRO24 representing the four serotypes of dengue, and pCRO28 representing Zika).
Surprisingly, given the extremely hydrophobic nature of the fusion loop (which features the residues W, F and L exposed at the tip of the E protein in close juxtaposition at its distal end in three dimensional space) in the case of dengue, all four representative serotypes tolerated substitution of two glycans (which are hydrophilic, and radically transform the topography of this part of the protein to an extent that mere amino-acid substitutions cannot) with no penalty to levels of expression (i.e., all expressed as well as the wild type sequence, in some cases markedly better). An objective had been set of ‘no less than wild type’ for levels of expression in order to ensure that the proteins were not misfolded which would have resulted in eradication from the endoplasmic reticulum via the ERAD channel for proteasomal degradation. Examples of the dengue serotype-1 sequence with a single glycan in the fusion loop were also made, but it did not express any better than wild type or the species with two glycans. In the case of Zika, attempts to generate variants with two glycosylation sites into the fusion loop (following the method established for dengue) were not successful, resulting in less secretion of the recombinant protein into the culture medium than for wild type.
In the case of the Zika E-protein exodomain we therefore explored the generation of variants with a single glycan at various sites in the fusion loop. Substitution of the tryptophan (W101), as for one of the dengue sequons, with an asparagine (the N of the sequon at 101 in place of W), resulted in a level of expression of the construct that was less than for wild type. Likewise, insertion of a glycan at F108 (i.e. the N of the sequon at 108, in place of F), resulted in a level of expression of the construct that was less than for wild type. We concluded that the Zika fusion loop was less tolerant to glycan insertion, and sought a more conservative way to allow it.
Having established, in the case of Zika, that neither the W101 nor the following F of the fusion loop could be replaced with the N of an N-linked glycosylation sequon, an alternative strategy was developed, which was not modeled on the approach taken for dengue. We sought to place a single glycan as near as possible to the end of the fusion loop (based on the 3D structure PDB 5IRE). Rather than go through the process of systematically making and testing the hundreds of possible variants that might allow glycan insertion (which would have been arduous by gene synthesis or by library technologies), we contrived a hypothetical solution and tested it. We contrived to straddle the W at the apex of the fusion loop with an N-linked glycosylation sequon. However, we reasoned that may have been infeasible by insertion of the classical NXS/T sequon, because W is not tolerated at the X position of a sequon. However, although W is not tolerated in the ‘X’ position in the centre of a sequon, H (histidine, a relatively conserved replacement for W, having a hydrophobic-aromatic/cationic dual character) can be tolerated in the X-position. We therefore substituted the 100 position with an N, used a H in place of the W for the X-position, and used a T (which we find works better with H than S), to make a single sequon that read ‘NHT’ (i.e. residues 100, 101, 102, using the E-protein numbering convention rather than the polyprotein numbering convention). The resulting protein, made from plasmid pCRO28, was found to express as well as wild type, and gave greater yield on purification than wild type, indicating no impediment to expression. The other variants of Zika that we explored gave rise to low level or no secreted protein in the expression systems used.
Glycan compositional analysis (GlycoThera, Germany) was performed on two of the selected proteins from Example 2, the dengue-2 serotype product of pCRO22 (representative of the selected dengue constructs that were all designed to carry two glycans in the fusion loop) and that of Zika (the product of pCRO28, designed to carry one glycan in the fusion loop) obtained from transfections of HEK 293.
The results of SDS-PAGE analysis of dengue and Zika samples prior to and after digestion with polypeptide N-glycosidase F (PNGase, Prozyme Inc.) are shown in
Glycans were released from the hyperglycosylated protein products and quantified by high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) and normal-phase HPLC with fluorescence detection of 2-AB-labelled N-glycans, along with specific exoglycosidase treatment (
Quantitative HPAEC-PAD analysis of native oligosaccharides was performed on an ICS 5000+ ion chromatography system of the Thermo Fisher Scientific Inc. (Waltham, Mass., USA; GlycoThera device-ID: HPAEC-7) using high resolution CarboPac PA200 columns. Injection of appropriate oligosaccharide reference standards was included in the analytical sequence.
N-glycans were detected via electrochemical detection. The data were collected and the chromatograms were acquired by using Chromeleon Chromatography Management System Version 6.8. Native N-glycans were analyzed via HPAEC-PAD revealing mainly neutral, monosialylated, disialylated and trisialylated oligosaccharides in both preparations according to GlycoThera's reference oligosaccharide standards. (
Desialylated N-glycans were analyzed via NP-HPLC after 2-AB labelling revealing predominantly complex-type N-glycans with significant permutational diversity, having proximal α 1,6-linked fucose in both samples (CV94=dengue-2, and CV95=Zika) according to GlycoThera's reference oligosaccharide standards. HPAEC-PAD mapping of native N-glycans released from dengue and Zika preparations CV94 (dengue 2 pCRO22 protein) and CV95 (pCRO28 protein) Zika (as shown in Table 2) revealed the presence of predominantly neutral (16.9% and 17.0%, respectively), monosialylated (30.7% and 36.9%, respectively), disialylated (26.6% and 32.0%, respectively) and trisialylated (15.0% and 8.4%, respectively) oligosaccharides in both samples. Significant amounts of tetrasialylated N-glycans (9.5% and 5.1%, respectively) as well as low proportions of pentasialylated/sulphated oligosaccharides (1.3% and 0.6%, respectively) were found in dengue and Zika samples CV94 and CV95; phosphorylated N-glycan structures such as oligomannosidic Man5-6GlcNAc2 glycan chains with one phosphate residue were not detected in either of the samples analyzed.
Site Occupancy Analysis of the Glycans:
Site occupancy was determined by LC-MS measurement of tryptic peptides. The analysis was based on the LC-MS measurement of tryptic or Endo Lys-C generated peptides liberated from proteins de-N-glycosylated enzymatically by PNGase F. Since PNGaseF is a glycoamidase, the asparagine (N) becomes converted to an aspartic acid residue (D). Quantification was done by creation of extracted ion chromatograms (EICs). The EICs were generated using the theoretical m/z values of differently charged target peptides within a mass window of +/−m/z of 0.01. In order to compare the peptide intensity with the specifically modified counterpart generated by de-N-glycosylation, the area of the peak of the EIC was used. The ratio/extent of modification was then calculated as follows: extent of modification=[area under EIC of modified peptide]/([area under EIC of modified peptide]+[area under EIC of unmodified peptide]).
Sequence numbering is by protein rather than the polyprotein sequence numbering convention, with W101 (at the very tip of the fusion loop) as a useful reference point. Sites are numbered according to their appearance in the linear sequence starting at the N-terminus, such that in dengue (pCRO22, GlycoThera sample number CV94) there were two additional sequons comprising sites 2 and 3. The Occupancy of the natural WT N-glycosylation sites was confirmed to be 100% and 99% for site 1 and site 4, respectively. The added N-glycosylation sites 2 and 3 (in the fusion loop) are located on one tryptic peptide (T15) and the occupancy was 38% (both sites) and additional 51% where only one of the two sites were N-glycosylated. In all 89% of the fusion loops had at least one glycan.
In the case of Zika, the occupancy of the N-glycosylation sites was confirmed to be 99.5% and 100% for the added ‘site1’ (residue 100, fusion loop) and site 2 (residue 154 the glycan naturally present), respectively. Site occupancy of the programmed glycosylation sequons was deduced from PNGase digestion and its effects on the mass of tryptic peptide fragments (whereby the amide NH2 group of the asparagine side chain is lost and converted to a hydroxyl group). (In the following sequences programmed sequons are in bold). In the hyperglycosylated dengue 2 exodomain the relevant tryptic peptide was T15, i.e., the 15th tryptic peptide (GN101GSGCGLN108GSGGIVTCAMFTCK122 (SEQ ID NO: 35)—containing the substituted N residues at 101 and 108. In the hyperglycosylated Zika exodomain (with a single introduced glycosylation sequon ‘NHT’) the relevant peptide was T10 (N100HTNGCGLFGK110 (SEQ ID NO: 36)).
These findings of efficient introduction of large and complex glycans into the fusion loop of dengue and Zika exodomain proteins strengthened our expectation that these proteins would neither bind to the fusion loop, nor elicit fusion-loop antibodies, giving confidence that B-cells or antibodies capable of recognising the wild type versions of the fusion loop would not engage with the glycosylated forms of the invention. This scenario is markedly different from mere introduction of mutations into the fusion loop, because by imposing one or more large additional glycan structures into the fusion loop, the resulting variant fusion loop cannot bind antibodies or B-cell receptors or generate fusion loop antibodies reactive with the wild type versions of the fusion loop. This was fully confirmed in later examples. This strategy may also be contrasted to deleting domains I and II from the structure of the protein, as these domains also contribute neutralising epitopes and T-cell epitopes useful for anamnestic immune responses upon encounter with flaviviruses in the wild, while pre-conditioning the immune system in such a way as to avoid the dangerous dominance of the fusion loop in immune responses to natural virus infections or to other vaccines.
(collectively, 89% of molecules have a glycan or two in the fusion loop. N101 replaced W101 of the WT sequence; N108 replaced F108 of the wild type sequence)
(99.5% of molecules have a single glycan in the fusion loop; N100 replaced G100 of the WT sequence)
Female Balb-c mice were immunized with PBS (negative control) and various dengue and Zika formulations of the hyperglycosylated exodomain proteins on Alhydrogel, alone (Zika mono) and in combination (Penta-) and as naked DNA (DNA). Alhydrogel formulations of proteins were injected subcutaneously (s.c.) in a total volume of 200 ul and naked DNA (comprising plasmids pCRO21, pCRO22, pCRO23 and pCRO24 of dengue plus pCRO28 representing Zika) was injected intramuscularly (i.m.) in a total volume of 50 ul for pentavalent DNA (representing 5 micrograms of each plasmid immunogen). Pentavalent protein combinations contained 5 ug amounts per dose of each hyperglycosylated exodomain, and monovalent (Zika) contained 10 ug per dose. Mice were dosed three times, once at each of day 0, day 14 and day 21. The legend at the bottom right of
The Balb-c Mice were immunized with DNA and protein representations of the glycoengineered exodomains and with the corresponding VLPs (i.e. VLPs representing the wild type sequences) from The Native Antigen Company Ltd, Oxford, UK (with no extra glycans, and exposed fusion loops) as positive control. These VLPs (see Table 8, used as both immunogens and also as test antigens in the ELISA tests of
There was little antibody response to naked DNA representing the five exodomains—as expected in the absence of delivery assistance from liposomal formulation, gene-gun or electroporation technology. Antibody responses to naked DNA were evident against dengue 1, 2 and 3 native VLPs, and not against Zika and dengue 4 VLPs. However these results served to demonstrate the potential utility of these DNA encoded antigens (all of them) with appropriate delivery systems. The assay is naturally more sensitive to detect immune responses to VLPs, due to the presence of additional epitopes (noted above), such that, as expected, antibody responses to the VLP antigens were uniform and very strong in the VLP-immunised ‘Group 4’. However, so too were responses to the novel glycoengineered exodomain proteins of the present invention, which gave strong, balanced immune responses against all five components (dengue serotypes 1,2,3 and 4 plus Zika) with the pentavalent immunogen formulation. Responses were uniformly high to the exodomain immunogens (pentavalent protein and monovalent Zika) and there were no non-responders. Also, the response to Zika in the monovalent-Zika-hyperglycosylated-exodomain-immunized group (10 μg dose) was modestly higher than that in the pentavalent protein group where the same exodomain was used at half the dose. This finding indicates a favorable lack of competition among the serotypes in the generation of type specific immune responses (this is a known problem with live attenuated flavivirus vaccine approaches, such as Dengvaxia, where immune responses to dengue serotype 2 are problematically low).
For direct ELISA (
Antibody responses were calibrated against fusion loop antibody 4G2 (The Native Antigen Company Ltd, Oxford) with dengue VLP representing serotype 2 on the solid phase at 0.5 micrograms per ml coating concentration. Units of antibody measurement “IgG antibody titre” are micrograms per ml 4G2-equivalent in undiluted serum, determined by interpolation of the standard curve using a four-component polynomial regression fit (AssayFit, IVD Tools). At day 42, antibody responses reached 104-105 for the hyperglycosylated exodomain immunogens (a notional 10 mg per ml-100 mg per ml in neat serum). These concentrations (taken literally) are unattainably high since the IgG concentration of mouse serum is only 2-5 mg per ml, and probably reflect the higher affinity or avidity of the antibodies generated compared to the antibody, 4G2, used for standardization, or may reflect better epitope exposure (4G2's fusion loop epitope being semi-crytpic in the structure of VLPs and virions). Nevertheless the 4G2 calibration serves a useful purpose allowing the assay to be run from time to time, controlling for such variables as batch to batch variation in the conjugate—(an anti-IgG-Fc horseradish peroxidase conjugate made from polyclonal antibodies which vary by batch). This is more reliable than quoting antibody ‘titres’ based on a threshold absorbance value which are very conjugate-batch and antigen-batch dependent, and may vary further among conjugates sourced by different manufacturers.
A further aspect of these observations is that the antibodies generated are of the IgG class demonstrating class-switching (even at day 14) from IgM, for all of the protein immunogens. This is an essential component of the B-cell memory response, important for the development of vaccines. A further aspect of these findings is that the antibodies generated by exodomain protein immunogens (and to some extent the DNA immunogens) strongly recognize the native form of the VLP antigens, which also lack His tags, ruling out the possibility of false positives due to anti-His-tag responses. This proves that both the dengue and Zika exodomain materials represent native epitopes of the exodomain proteins that are immunogenic in generating anti-viral (VLP) antibodies. These results suggest that other nucleic acid encoded forms of the hyperglycosylated exodomain species, e.g., liposomal RNA or lipoplex RNA, would also generate desirable antibody responses against virions (VLPs) and viruses.
There was specificity in the immune response to the Zika monovalent hyperglycosylated exodomain, which generated higher antibody titres against the homologous Zika VLP than to other VLPs, despite the known cross-reactivity of these various viruses with antibodies. This is a favourable result since type-specific anti-Zika antibodies are known to have better neutralizing activity generally than dengue-cross-reactive ones. Also, as seen in the antibody-responses to the Zika-monovalent hyperglycosylated exodomain at the later time points (after two or three doses), there was a degree of cross-reactivity against dengue strains that developed over time, raising the potential for generation of beneficial cross-reactive neutralizing responses, excluding the fusion loop epitope (which was not recognized by antibodies generated by hyperglycosylated exodomain species as demonstrated in the data that follows in later examples).
An ELISA test (of
Unless otherwise specified, conditions were the same as for the ELISA test of Example 4 and
Antigens were as follows: wild type dengue exodomains representing dengue serotypes 2 and 4 were from The Native Antigen Company (DENV2-ENV, DENV4-ENV); ‘HX’ designated exodomains (hyperglycosylated exodomains) were the selected set of Excivion exodomains of the present disclosure (pCRO21-24 for dengue, pCRO28 for Zika). Prospec Zika was a non-glycosylated bacterial exodomain from Prospec of Israel (zkv-007-a), and Aalto Zika was an insect (Sf9 cell) derived Zika exodomain (AZ6312-Lot3909). Mouse monoclonal antibodies against Zika virus exodomain were as follows: Aalto Bioreagents AZ1176-0302156-Lot3889; Z48 and Z67 were neutralizing antibodies described by Zhao et al, Cell 2016 (The Native Antigen Company ZV67 MAB12125 and ZV48 MAB12124). Antibody 4G2 is an anti-dengue-serotype-2 antibody recognizing the fusion loop (The Native Antigen Company AbFLAVENV-4G2).
The data of
The data of
A further aspect of the data of
An ELISA test was established to measure the binding of polyclonal antibodies against the fusion loop (represented in this example by dengue serotype-3 VLP on solid phase ELISA plates).
A competition ELISA was set up using biotinylated 4G2 (Integrated Biotherapeutics) which was detected using streptavidin-horseradish peroxidase conjugate. Dengue serotype 3 VLP (The Native Antigen Company) which reacts with 4G2 slightly better than the immunizing serotype dengue-2 VLP was used as antigen coated at 0.5 ug per ml on the solid phase. Pooled sera (from the groups of
In this assay (
In the case of Zika, inhibition was detectable only at the highest concentration tested, indicating a >1000 fold advantage in avoidance of fusion loop antibodies compared to VLP immunogens, if this single point at 1/10 serum dilution is (for the sake of argument) deemed to be significant.
The data of
Serum pools from Example 4 were tested for their ability to neutralize dengue serotype 2 and Zika viruses using Vero cells in plaque reduction neutralization tests (PRNT).
In the case of dengue, the dengue serotype 2 strain used to infect the Vero cells (D2Y98P) was a different serotype-2 strain (non-homologous) from the sequence of the immunizing dengue 2 strain of the VLPs and exodomains. In the groups expected (from Example 4) to generate dengue neutralizing antibodies (namely pentavalent protein and pentavalent VLPs, Groups 2 & 4) there was potent neutralization of the ‘off target’ dengue test virus. In the case of Zika there was significant (albeit partial) neutralization as expected from the results of Example 4, in groups shown to contain antibodies that recognized native Zika VLPs (namely pentavalent protein and pentavalent VLPs, Groups 2, 3 & 4). Due to limitations on sample volume, the maximum concentration of serum that was tested was 1/50, such that in interpreting these results this factor needs to be taken into consideration (i.e. that there would be higher neutralizing capability in the blood of the immunized animals).
PRNT Assay was performed as follows. Five mouse serum samples were pooled by taking an equal volume of individual samples in each group (sample description in next slide) and were then tested against ZIKV and DENV, respectively. Twelve two-fold serial dilutions of each serum sample in duplicates starting at 1:50 were prepared for the two-hour inoculation with virus. The serum-virus mix was then added to Vero cells seeded in 24-well culture plates and incubated at 37° C. in a humidified 5% CO2 atmosphere. The Vero cells were fixed on 3 days post incubation (dpi) for ZIKV PRNT and 4 dpi for DENV PRNT. Viral plaque was determined by crystal violet staining.
Potent inhibition of infection by dengue was observed in the group immunized with hyperglycosylated exodomain proteins of the present disclosure (Penta-prot). Zika immunized animals generated antibodies that did not prevent dengue infection of Vero cells, illustrating the type-specific nature of antibodies generated by these novel immunogens. These Zika antibodies (from the Zika monovalent group and from the pentavalent proteins group) were significantly protective of infection of Vero cells by Zika virus. As expected, PBS-sham-immunised animals did not give rise to protective antibodies, nor did pentavalent DNA administered intramuscularly. This latter result may have been due to the low concentrations of antibodies generated by naked DNA, as expected from intramuscular injection (as distinct from gene-gun or electroporation strategies, or strategies incorporating encoded proteins as molecular adjuvants).
The results of Example 6 (generation of neutralizing antibodies) combined with those of Example 5 (lack of recognition by or generation of fusion loop antibodies) by the hyperglycosylated Exodomain proteins of the invention strongly suggest that these proteins can form the basis of a protective vaccine for dengue or Zika viruses (or, in combination, for both viruses) without the generation of fusion loop antibodies, which are particularly implicated in antibody-dependent enhancement of infection.
The ELISA reactivity of antibodies in a dengue convalescent serum with immobilized Zika and dengue wild-type (WT) and hyperglycosylated (HX) exodomain proteins oriented on the solid phase by capture with a rabbit anti-His-tag monoclonal antibody (
The ELISA reactivity of antibodies in a Zika convalescent serum with immobilized Zika and Dengue wild-type (WT) and hyperglycosylated (HX) exodomain proteins (
The results show that:
1) the HX Zika antigen of the invention is not susceptible to the off-target recognition of WT Zika exodomain by the convalescent dengue serum.
2) The off-target recognition of WT Zika exodomain (Aalto) by dengue serum is a fusion-loop directed phenomenon because it is abolished by 4G2 (anti-fusion loop monoclonal antibody) in solution phase at a concentration that causes 80% inhibition against VLPs (10 micrograms per ml). (The antigen on the solid phase in this instance is exodomain rather than VLP).
3) The ‘Zika’ convalescent serum does not recognize any of three Zika exodomains, but it strongly recognizes WT dengue 2 and WT dengue 4. In the Example 6 the HX Zika antigen of the invention and Aalto's Zika exodomains exhibit reaction with conformation-dependent anti-Zika neutralising antibodies). This demonstrates that this particular Zika serum (positive for Zika plaque neutralisation and Zika NS1 antibodies) is from a subject also exposed to another flavivirus. Because the Zika convalescent serum (unlike the dengue convalescent serum) does not recognize the fusion-loop-cloaked exodomains, it can be concluded that this other flavivirus is not dengue.
4) The off-target recognition of WT dengue-2 and dengue-4 exodomains by the human Zika convalescent serum is not seen with the HX-cloaked dengue exodomains of the invention. This suggests that it is fusion loop directed and would show false positive in other flavivirus diagnostic tests that do not use glycan-cloaked proteins in accordance with the invention.
5) The off-target recognition of WT dengue-2 and dengue-4 exodomains by the human Zika convalescent serum is blocked completely by 4G2 showing that it is a fusion loop directed phenomenon.
6) The dengue convalescent serum recognizes WT 2 & 4 indiscriminately, but clearly prefers the d2 exodomain out of the set of 4. This demonstrates that the fusion loop antigens of the invention have superior selectivity (compared to their wild type equiavalent forms) to discriminate between dengue serotypes, due to the glycan cloaking of the fusion loop.
Examples of Diagnostic Use
Various formats of lateral flow test for the detection of antibodies are possible. A widely-used design for the detection of antibodies against viral antigens is to place a line (or spot) of antigen ‘directly’ on a strip of nitrocellulose, and then allow blood/serum/plasma or oral fluid to seep up the strip by capillary action, hydrating a dried nanoparticulate gold reagent, wherein the particles are coated with an antibody (e.g. anti-IgG, anti-IgM), typically followed by a wash buffer. As the reagents proceed along the nitrocellulose strip, specific antibodies are arrested by the antigen, and some of the gold particles are likewise arrested by binding to specific antibodies immobilised on the antigen, excess particles proceeding beyond the observation window—resulting in a brown or pink line in the observation window of the test cassette indicating the presence of specific antibody. In some embodiments of this format, a two-port system is used, one port for the sample application (proximal to the antigen line), and a second port, more distal, for the application of the diluent. Additionally, a control line (or spot) may be used to verify the fluidic performance the test and the functionality of the gold conjugate. This may be formed of a line of purified antibody or of human serum containing the antibody type appropriate to the anti-Ig conjugated nanoparticulate gold reagent. In preliminary studies with this format, we were successful in demonstrating the diagnostic utility of wild type, and the superior specificity of HX versions of the four dengue envelope proteins and Zika. In addition, aware of the phenomenon of surface denaturation of proteins (Sen, Yamaguchi, & Tahara, 2008) which is more common in our experience with nitrocellulose (viz. lateral flow) as a solid phase compared to polystyrene (viz. ELISA) we were influenced in the current lateral flow design by our earlier ELISA studies (see above), to employ indirect immobilisation of the HX envelope proteins. Indirect immobilisation had proven more sensitive in ELISA studies than direct immobilisation of the antigen on the solid phase. To this end we explored the use of anti-His-tag antibodies recognizing the C-terminal hexahistidine tag on the HX proteins described herein as an indirect means to attach the antigens to the solid phase (Example 9).
Test Composition (common to both formats A and B of this Example 9)
Absorbent pad—Whatman CF5 (22 mm)
Nitrocellulose—MDI 15μ SS12 (25 mm)
Sample pad—Ahlstrom 8964 (20 mm)—Sample pad can be treated with a red blood cell arresting agent. This can be anti-glycophorin A (AGA) at a suitable concentration (suitable concentrations are known in the art). Pad is soaked with anti-glycophorin A and dried for 24 hrs at 25° C. in low humidity.
Conjugate pad—Ahlstrom 8815 (Format (A) 20 mm, Format (B) 45 mm)—Conjugate pad is soaked with 0.1M Tris-HCl pH8.0 containing 0.25% Tween 20 and 0.5% bovine serum albumin (BSA) and dried overnight at 37° C. followed by 24 hrs at 25° C. in low humidity
Cassette—Kanani Biologicals, India Gold conjugate (detector) 40 nm colloidal gold (BBI Solutions of Cardiff, Wales, UK) conjugated to anti-human-IgG at 100 units per test on the conjugate pad, dried for 24 hrs at 25° C. in low humidity. Advantageously the conjugate is made from affinity-purified IgG (e.g., from goat or rabbit) or a suitable monoclonal antibody such as a pan-subclass-reactive IgG antibody or a monoclonal anti-IgM antibody. Such antibodies are known in the art and are available from Sigma, AbCAM or Merck-Millipore. Gold conjugates (where applicable) were applied in 2 mM tetrasodium borate, 10% sucrose, 0.95% sodium azide pH 9.0.
Running Buffer was prepared from Oxoid PBS pH 7.3 tablets using deionised water (NaCl 0.137M, KCl 0.003M, disodium hydrogen phosphate 0.008M, potassium dihydrogen phosphate 0.0015M), containing additionally 0.5% Tween-20 and (where indicated) 0.5% PEG 20K (average Mn 20,000) CAS Number 25322-68-3.
Antigens
The antigens used were as follows: (in ‘direct’ formats, antigen was applied directly to the nitrocellulose strip in Oxoid-PBS without additives and dried as above; in ‘indirect’ formats, where antigen was applied to the sample application pad, antigens were applied in Oxoid-PBS without additives and dried as above).
Dengue HX antigen D1: Hyperglycosylated exodomain D1 (from pCRO21) (SEQ ID NO: 24)
Dengue HX antigen D2: Hyperglycosylated exodomain D2 (from pCRO22) (SEQ ID NO: 25)
Dengue HX antigen D3: Hyperglycosylated exodomain D3 (from pCRO23) (SEQ ID NO: 26)
Dengue HX antigen D4 Hyperglycosylated exodomain D4 (from pCRO24) (SEQ ID NO: 27)
Zika HX antigen: Hyperglycosylated exodomain Zika (from pCRO28) (SEQ ID NO: 28)
The alignment of the E-protein fusion loop amino acids 98-110 of a group of wild-type sequences of flaviviruses and recombinant analogue sequences of the invention is shown in Table 1, for each of the Dengue HX antigens D1-D4 the hyperglycosylated exodomain fusion loop sequence was DRGNGSGCGLNGS (SEQ ID NO: 2) for the Zika HX antigen the hyperglycosylated exodomain fusion loop sequence was RNHTNGCGLFGK (SEQ ID NO: 5) Dengue HX antigens D1, D2, D3 and D4 at 100 ng each antigen per test on the sample pad, dried for 24 hrs at 25° C. in low humidity.
Zika HX antigen at 50 ng per test on the sample pad, dried for 24 hrs at 25° C. in low humidity.
Nitrocellulose
Test Line—Mouse mab anti-6×His at 0.3 mg/mL (RT0266 Bioxcell)
Control Line—Human IgG at 0.75 mg/mL (Sigma-Aldrich-Merck 18640—IgG from human serum)
A: Single Port Format for Lateral Flow Detection of Antiviral Antibodies
As shown in
B: Two-Port Format for the Lateral-Flow Detection of Antiviral Antibodies
As shown in
A: Using the test format of Example 9A (two ports used) 15 ul of normal human plasma (spiked with a two-fold serial dilution of human dengue monoclonal neutralising antibodies DV78 and DV18 (Absolute Antibody Ltd., Oxford UK) in equal amounts, from 1 to 0.0078 ug) was added to the square sample pad port, and allowed to incubate at room temperature for 3 minutes. Next, 45 ul of running buffer was added to the sample port, and the device was run for a further 17 minutes. 90 ul of running buffer was then added to the round port to release the gold anti-human-IgG conjugate. The test was read after a total of run time of 40 minutes. This demonstration illustrates a sensitive capability of this test format to detect neutralising anti-dengue antibodies (DV78 and DV18 recognise epitopes in the DI and DII of dengue virus), 7.8 ng of specific antibody was detectable. In this particular example HX versions of dengue-2 and dengue-4 exodomains were used.
B: Results obtained in this format with control serum NM (not exposed to flavivirus infection or vaccination) and a dengue positive serum ‘2965’ are shown in
Example 10 demonstrates the utility of the HX dengue exodomains 2 and 4, in serodiagnosis of dengue in lateral flow format. It also demonstrates the advantage and utility of capturing the antigen (initially mobile) using an anti-tag antibody. A monoclonal anti-His-tag antibody was used, although there are other tag-antibody pairs that could be used (such as FLAGtag and StrepTag-II). (Such additional pairs can be used to multiplex the assay, whereby a second line of anti-tag would recognize a second tag on a second antigen involved in the test, a third a third etc.). By using a tag antibody for capture of the mobile antigen (and antigen-antibody-complexes and antigen-antibody-colloidal gold conjugate materials), a prozone effect is avoided. This contrasts to an alternative scenario where one might have used an antigen-specific (e.g. murine) monoclonal antibody for the purpose of arrest at the test line or spot. In this latter alternative scenario, the test is defeated by the presence (if present) of an antibody in the test sample against the same or overlapping epitope as is recognized by the antigen-specific capture antibody. (Such adverse competition for antigen binding can be mitigated by the use of multiple monoclonal or a polyclonal affinity purified antigen specific antibody, but cannot be eliminated by this means). In the Example 10 and elsewhere in other examples we have detected IgG antibodies by virtue of a colloidal gold conjugate with surface bound anti-human-IgG. However, it should be apparent that other antibody classes and subclasses can be detected in this way (e.g. IgM, IgA and subclasses of IgG) by use of appropriately specific antibodies on the colloidal gold reagent, and likewise by use of of corresponding pure antibodies comprising the control line (i.e. IgM for an IgM test, IgA for an IgA test, IgG for an IgG subclass-1 for an IgG subclass-1 test etc.). For tests designed to measure IgG, when using a monoclonal or affinity-purified anti-IgG colloidal gold conjugate, it is important to choose one that has a balanced reactivity across the IgG subclasses, in order to give a representative signal in the lateral flow tests. Such antibodies are known in the art, and are commercially available.
The antigens used in serodiagnostic tests are sometimes more expensive than antibodies when antibodies are used in such tests. This is particularly important in lateral flow tests targeting mass markets. (This is a feature of improvements in recent years in the facile production of monoclonal antibodies in particular, e.g. by transient expression in HEK cells, where the antibodies can be produced at high yield, e.g. several grams per litre of culture). Therefore, an additional advantage of exploiting anti-tag antibodies for indirect antigen-capture in lateral flow tests is that the test is made more economic with respect to antigen costs in manufacturing. The economic advantage of this approach is further enhanced by the stronger recognition of indirectly-immobilised antigen (via anti-tag antibody), compared to that deposited directly on the lateral flow strip (e.g. a nitrocellulose strip).
A further feature of Example 10 is demonstration that the HX class of flavivirus exodomain antigens (having glycans in the fusion loop) have intact neutralising epitopes throughout the molecule (except, as noted above, for the fusion loop—in which the epitope is glycan-cloaked).
Thus as described above it was demonstrated for Zika HX exodomain that immobilisation via its His-tag allowed functional display of two neutralising epitopes, namely those recognised by murine monoclonal antibodies ZV48 and ZV67 (human antibodies from Absolute Antibody Ltd., Oxford). Example 10 demonstrates that HX versions of dengue-2 and dengue-4 exodomains functionally display epitopes recognized by DV78 and DV18, which are located in domains [I and II]. These data and arguments strongly support the use of the HX antigens in serodiagnostic tests, in particular point-of-care serodiagnostic tests, for the detection and measurement of neutralising antibodies. Antibodies against the fusion loop, which dominate the antibody response, are poorly neutralising and rapidly decline to concentrations that enhance infection. The point-of-care diagnostics described here for dengue and Zika can therefore be used to measure vaccine performance, e.g. to determine whether an additional dose may be needed, or how well a vaccine is performing in a subject already exposed to another flavivirus by reason of infection or vaccination (e.g. with yellow fever or dengue vaccines). They may also be combined (dengue and Zika, by using different tag/anti-tag pairs, e.g. His-tagged dengue antigen being captured by an anti-His-tag line, and FLAG-tagged Zika antigen being captured by an anti-FLAG antibody).
Results: 1&2 verify the performance of the conjugate in expected rank of intensity; 3 showed a very faint signal (not visible in the photograph), likewise 4; 5&6, controls for non-specific binding of reactants, were negative; 7&8 (negative and positive test sera) gave the desired result, i.e. negative in 7, positive in 8. These findings demonstrate (in the case of directly applied immobile antigen) the value of an intervening wash, because signal was weak (not evident on photograph) when no wash was used. Moreover, relatively large amounts of antigen were required, even of the strongly reactive fusion-loop-intact forms, in this test configuration. (Where reagents or solutions are not defined in this Example, they are the same as in Example 9).
Example 11 demonstrates the utility of direct immobilisation by spotting of exodomain antigens on nitrocellulose, as distinct from having them in the mobile phase of the test (which is the format in all other Examples in this application). Direct spotting/immobilisation of antigen on nitrocellulose (as distinct from capture of mobile phase antigen via anti-tag) was effective, but less sensitive than indirect immobilisation of mobile antigen via anti-tag antibody. In Example 11 wild type antigens are used (dengue-2 and dengue-4) but it is clear that HX versions of flavivirus antigens could be used in this format. In the format of Example 11 it would be possible to improve sensitivity by use of luminescent detection, as described (Laing, 1986), or using fluorophore-labelled anti-human-Ig, with a fluorophore such as a quantum dot or a lanthanide which have high quantum efficiency and low photobleaching, as well as a favourable Stokes shift (particularly in the case of lanthanides such as Europium chelates). Europium chelates are also amenable to time resolved fluorescence which is more sensitive because the fluorescence measurement is made once the light source is turned off, and does not contribute noise to the signal. Clearly these sensitive luminescent and fluorescent methods could be applied to other Examples given in this patent application, if desirable or necessary.
As shown in
These results show a strong off-target recognition of wild type dengue antigens by the Zika convalescent serum which is not exhibited by the tests conducted with the HX versions of the same antigens, which nevertheless correctly identify the dengue positive human serum. The macaque ‘G16’ was bred in captivity and cannot have been exposed to dengue. The strong ‘quasi specific’ recognition of the WT dengue antigens by the macaque serum is a true antigen-specific reaction, explained by reaction of anti-Zika antibodies in the macaque serum with the dengue fusion loop, which is effectively cloaked (masked) by glycosylation in the case of the HX antigens. Results by test are 1 negative; 2 negative; 3 positive; 4 weakly positive (much less than false positive in 3); 5, positive; 6 positive; 7 weak positive; 8 negative. 100 ng of each antigen (200 ng total) was used in each test. ‘8’, using HX versions of the dengue antigens, was negative as expected (desirable result) whereas ‘7’ gave rise to a weak non-specific binding reaction of this non-flavivirus exposed subject (VM) to the wild-type version of the antigen (undesirable result).
Example 12 demonstrates the superior performance of dengue-HX antigens (combined dengue-2 and dengue-4 HX antigens) compared to their wild type equivalents, the wild type versions showing strong off-target recognition by anti-Zika antibodies. In this example a captive-bred macaque was used. This obviates the uncertainty, in the case of human sera, that a person may also have been exposed to dengue as well as Zika, because there is high co-endemicity of Zika and dengue such that the majority of Zika-exposed human subjects will also have been exposed to dengue.
HX versions of the four dengue proteins were made in either human embryonic kidney cells or in insect (Tni) cells using a secretory-type baculovirus vector expression system (Flash-Bac Ultra) according to manufacturer's instructions.
In this example the reactivity of HX versions of the complete set of dengue antigens 1, 2, 3 &4 (as a mixture) and of the HX Zika antigen are compared for a dengue positive serum+D=serum 2965, and a Zika positive serum+Z (G16 macaque Zika convalescent serum, and a control human serum positive for neither.
Example 14 demonstrates utility and specificity of HX Zika and dengue antigens in lateral flow testing compared to wild type Zika antigen. In this example insect-cell expressed dengue and Zika antigens were used. The extremely strong reaction seen with the HX version of the insect-expressed Zika antigen can be traded off with respect to specificity, by using less antigen in the test or less anti-His-tag antibody in the test line, in order to render the weak off-target recognition of dengue HX antigens undetectable (i.e. invisible).
Example 15 demonstrates the performance of various configurations of the test with whole blood, all based on the format of Example 9A (using a single port). For a point-of-care test it is most useful and convenient if it can be conducted using blood without the need for plasma or serum separation. First it was investigated whether it was advantageous to use an anti-glycophorin antibody in the sample pad to absorb-out red blood cells preventing them from running on the nitrocellulose and complicating the reading of the device. It was found that the anti-glycophorin antibody reduced the signal of the test to an unacceptable degree. However, it is noted that it may nevertheless be advantageous to include a lesser amount of such an antibody in the sample pad, in order to absorb out heterophile antibodies (human anti-mouse-IgG antibodies) that occasionally give false positives in tests using mouse monoclonal antibodies. In the format of Example 15, without anti-glycophorin antibody, specific reactivity of dengue positive samples was demonstrated without problems of false positivity due to yellow fever vaccination in five subjects. Also in this format, there was sensitive detection of dengue neutralising antibodies D18 and D78.
A: The test requirement for anti-glycophorin A (AGA) in the dengue test was investigated. Tests were run according to Example 9A, but also including 0.5% PEG20K in the running buffer. Samples were made from serum mixed with an equal volume of centrifuge-sedimented red cell fraction from whole normal blood. 10 ul samples of blood were used. 100 ng of each HX dengue antigen was used and 100 units of anti-human-IgG gold conjugate.
B: Tests were run as in 15A above including testing of serum JF, a yellow-fever vaccine. No AGA was used. There was minimal breakthrough of red blood cells into the test window (
C: Tests were run as above in 15A and 15B, with 5 ul of negative sample which had been spiked with various levels of D18/D78 antibodies (Absolute Antibody Ltd., Oxford) in equal mixture. 100 ul of running buffer was used as chase. All tests used 100 ng of each dengue insect derived HX antigen and 100 units of anti-human-IgG gold conjugate. The results are shown in
D: Blood samples were made up by replacing the plasma of a normal blood sample from which red blood cells were sedimented, with various plasmas (negative, positive and yellow-fever vaccinated) and resuspended. 100 ul of running buffer was used as chase. 100 ng of each of the four insect-derived dengue HX antigens were used, plus 100 units of anti-hu-IgG gold conjugate. “+” is the positive control used in the various previous test examples above. A very favourable profile of positivity was observed—all positive samples gave a positive test line, none of the yellow fever vaccinated subjects or the negative control subject, gave a positive test line (
HX proteins were initially produced in HEK cells, however it was difficult to achieve efficient expression of some of the proteins. To seek to improve protein expression, HX proteins were produced in Tni insect cells in the ‘Flash-Bac-Ultra’ baculovirus system of Oxford Expression Technologies Ltd. According to the manufacturer's intructions. Protein purification was performed using Strep-TactinXT® Superflow® system(IBA Lifesciences), according to the manufacturer's instructions.
The results of expression of Zika HX-Strep-tag-II from insect (Tni) cells are shown in
In Tni cells, an approximately 10-fold improvement in productivity (relative to expression in HEK or CHO) was achieved for the ‘difficult-to produce’ dengue-1-HX and Zika-HX antigens, which now expressed at levels approaching those of the most abundantly-expressed antigens such as dengue-2-HX. High levels of expression have been achieved subsequently in CHO cells, by optimising signal sequence and by cell-line development.
As it was possible that the HEK-expressed antigens might differ from the baculovirus-expressed forms (notably these cell lines produce different glycan structures), bridging studies were performed with the lateral flow tests to determine if these differences influenced the performance of the test with human sera, or whether there were differences in reactivity in ELISA format with panels of monoclonal antibodies.
The antigenic integrity of insect (Tni) expressed HX forms of the dengue and Zika proteins (having one or more glycans in the fusion loop) was probed with a panel of murine anti-Zika monoclonal antibodies (some of which had known specificity and properties), and with 4G2—a flavivirus cross-reactive antibody which was raised against dengue-2 wild type and which recognizes the conserved fusion loop (Table A). This was done in order to ensure that the data generated earlier on the HEK-expressed proteins would translate effectively when the insect (Tni) expressed antigens were used in the LF tests, in place of the HEK-expressed proteins. These studies confirmed that the baculovirus-produced antigens were equivalent to the HEK-expressed forms in their recognition by by well-characterised murine monoclonal antibodies. Also, glycan analyses were conducted on the baculovirus-expressed proteins which confirmed that the fusion-loop planting efficiency of the glycans was the same for the baculovirus-expressed as was the case for the HEK-expressed proteins, although the glycans were, as expected, different between insect and HEK expression.
Comparison of Excivion dengue-LF prototype test of the invention to the SD BIOLINE Dengue DUO® (SDB DD) NS1 Ag and IgG/IgM ICT (SD Dengue Ag+Ab Duo) test was performed in a field study carried out in Pune, India.
The manufacturer's instructions for the use of SD BIOLINE Dengue DUO® (SDB DD) NS1 Ag and IgG/IgM ICT (Standard Diagnostics, Inc., Korea) were followed, and are described previously (Wang S M, Sekaran S D. Early diagnosis of dengue infection using a commercial Dengue Duo rapid test kit for the detection of NS1, IgM, and IgG. Am J Trop Med Hyg. 2010; 83(3):690-5).
The Excivion dengue-LF prototype test (without preabsorption).
The Excivion dengue-LF test was used to perform 47 tests on the sera tested previously with SD Dengue LF test (
Excivion-Omega dengue-LF test had “increased sensitivity for IgG”. (statistical analysis: Wilcoxon Matched Pairs Test N 47 T 0. Z 5.23162 p-level 1.6803E-7).
The performance of early test prototypes (no preabsorption) was assessed first with panels of sera available from commercial and academic sources. Tests were performed using dengue and Zika serum and plasma panels collected in endemic territories or from returning traveler populations including Thailand, the Dominican Republic, USA and UK-Trinidad, collected before and ‘after’ the advent of Zika as a pandemic disease, although it became apparent that Zika had arrived in Brazil in 2013 (earlier than previously thought). The test was also demonstrated to work with whole blood, its intended use.
Tests were also performed on NIBSC sera, noted above, from non-human primates that had been exposed only to Zika. The specificity seen in earlier ELISA tests with the NIBSC primate sera, was replicated in the LF format Zika test. These primate Zika-convalescent sera were positive in the Zika LF test, but negative in the dengue LF test, employing our proprietary ‘HX’ antigens. In contrast, the wild type Zika antigen (i.e. without cloaking of the fusion loop) gave rise to false-positives with dengue sera when used in LF-tests. These findings confirmed the superior diagnostic accuracy of our HX antigens over the wild-type (natural) equivalent proteins, in a lateral-flow (as well as ELISA) context.
When human sera were used in the LF tests, we found that the tests had excellent sensitivity (judged initially by spiking normal human sera with monoclonal anti-Zika and anti-dengue antibodies). However, we found that there was a significant false-positivity rate in both the Zika and dengue LF tests, employing our HX antigens, when endemic panels of sera were tested. We attributed these false positives to ‘mosaic’ epitopes representing short segments of identity in amino-acid sequence common between dengue and Zika viruses (and to a lesser extent with other human flaviviruses), less immune-dominant than the fusion-loop. We were able to obviate false signals due to yellow fever vaccination by increasing the ‘stringency’ of the test (via manipulating running-buffer composition), however this was not sufficient to abolish all false signals originating from (putative) dengue-only and Zika-only sera. We then embarked on a modified strategy of test design, which incorporated ‘off-target-pre-absorption’ to deal with remaining cross-reactions.
Preabsorption is a well-established technique which is sometimes used in serological analyses to prevent signals being generated by cross-reactive antibodies. In the case of flaviviruses this is difficult to achieve, but our advent of cloaking the fusion loop and obviating signals from the major cross-reactive site of this family of viruses makes it possible to exploit preabsorption to minimize signals from cross-reactive antibodies in LF tests for Zika and dengue. In the Zika test we have described Zika as the target antigen, and preabsorbing dengue antigen as the ‘off-target’ antigen. In the dengue test we have described (below) dengue 1,2,3,4 as the target antigen and Zika as the off-target antigen. Because the mosaic epitopes (defined above) were present in sequence alignments across Zika and all four strains of dengue, we used dengue-2-HX for preabsorption instead of all four dengue serotypes. In configuring tests with preabsorbing antigens it was necessary to construct versions of dengue-2-HX (SEQ ID NO: 2) and Zika-HX (SEQ ID NO: 5) that had either no tag or a different tag than a His tag, otherwise they would contribute signal in the test from cross-reactive antibodies. For this reason we constructed Strep-tag-II versions of the dengue-2-HX and Zika-HX antigens. HX versions were used for preabsorption because they were more productive than the wild type versions in recombinant expression. By configuring the tests in this way we were able to engineer the capture and blockade of antigen combining sites of cross-reactive antibodies with Strep-tag-II versions of the off-target antigens, whereby such antigen-antibody complexes were not captured by the anti-His-tag line and ran off the end of the observation window of the test without contributing signal.
As multiple-exposure (and cross-reactivity of antibodies) was expected to be commonplace in endemic areas, it was desirable to employ tests with off-target pre-absorption. Thus we set about determining what amounts of off-target antigen would be needed in each of the pair of LF devices (one Zika and one Dengue test making up the pair), and found that 1 ug per device was adequate for the dengue test, and 5 ug per device for the Zika test (these larger amounts of antigen compare to 200 ng amounts of the test antigen in each device). In the case of the Zika test, off-target preabsorption using all four dengue antigens in Strep-tag form was considered, but after comparison of the sequences and X-ray structures of the wild-type Zika and dengue antigens, it was concluded that use of dengue-2-HX-Strep-tag alone would be sufficient for off-target pre-absorption, because the ‘islands of identity’ in the dengue and Zika HX sequences (comprising residual cross-reactive ‘mosaic’ epitopes) were essentially the same across all four dengue serotypes, additionally the dengue-2-HX was easiest to produce. Pre-absorption with the panel sera with dengue-2-HX was successful and provided LF Zika devices, equipped with off-target pre-absorption built into the LF device for the field testing in Brazil (which were tested alongside the analogously configured dengue LF test devices).
In a particular embodiment, the Zika test contained non-His-tagged (Strep-tagged) dengue antigen, and (conversely) the dengue test contained Strep-tagged Zika antigen, the off-target antigens were included to bind to cross-reactive antibodies, but not result in gold-particle-arrest (visually detectable signal) on the test line (which was effected by an anti-His-tag antibody). Because the HX antigens were much easier to produce, the off-target antigens were made as HX (rather than wild-type) forms.
In conventional LF tests for this purpose, test line has an anti-Ig (anti-IgM or anti-IgG) and the gold conjugate is labelled with antigen, giving rise to arrest of gold particles by human antibodies that bridge the solid-phase anti-Ig of the test line with gold particles. In our preferred configuration, the test-line is monoclonal-anti-His-tag. Off-target antigens (for pre-absorption) are Strep-tagged (using StrepTag-II), allowing them to flow past the test line. Of a number of architectures tested, we found this configuration to be most sensitive, while simple and easy to manufacture. By use of monoclonal and recombinant reagents exclusively, it is ‘infinitely’ scalable.
A depiction of embodiments of the LF test devices is given in
In the devices used in the clinical field studies in Rio, the LF tests used 1 ug of Zika-HX and 5 ug of den-2-HX for off-target-pre-absorption of cross-reactive antibodies.
In field testing conducted in Rio de Janiero, the Zika test was found to be 100% sensitive (positive in 50 samples out of 50) in detecting cases of Zika confirmed by PCR (in samples from 2016) that were found to be positive in the blockade of binding assay ‘BOB’, a recently developed ‘second generation’ Zika ELISA assay based on blockade of NS1 monoclonal antibody binding, regarded as the most reliable laboratory-based test for Zika antibodies (Balmaseda A et al Proc Natl Acad Sci USA. 2017 Aug. 1; 114(31):8384-8389. doi: 10.1073/pnas.1704984114. Epub 2017 Jul. 17).
Using our tests all of these samples were strongly and unambiguously (++++) positive (whereas in the BOB test they exhibited a range of positivity). This result tallied with our own observations with panel sera in which our LF tests were found (in the case of dengue) to be very significantly more sensitive than Standard Diagnostics' LF test with respect to IgG detection. Our Zika test was also very good at detecting seroconversion (elevation of antibodies indicative of recent infection) in serial samples (p<0.01)
In the Rio field testing, the earliest sera we were able to obtain were from 2014, which was believed to be before significant Zika circulation had occurred in Rio, although it has recently been recognized that Zika was, in fact, circulating in Brazil in 2013, meaning that there could be some Zika cases in this sample set collected in 2014. We found a 4% positivity rate for Zika (calling any result >3 as positive) with our LF test in this group. However, two tests for Zika NS1 antigen put this figure even higher—at 50%. This result indicated that our Zika test is more specific than the commercial NS1-based Zika ELISA tests, which are false positive in 50% of cases. With respect to our dengue LF test, we found a positivity of 92% (46/50) in the 2014 samples, which tallies with an expected 90-95% dengue seropositivity in the general (Rio) population from various literature studies. The results of testing are shown in
We have developed a pair of cheap, simple point-of-care devices for detecting and distinguishing prior dengue or Zika infection, the devices are convenient, do not require a laboratory environment for their performance. The performance of the tests is better than previous commercial LFs and ELISAs and they will find many uses in the diagnosis and monitoring of these infections, including their demonstrated use as a ‘companion diagnostic’ with potential to increase the safety of use of vaccines.
| Number | Date | Country | Kind |
|---|---|---|---|
| 1719423.4 | Nov 2017 | GB | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2018/063422 | 5/22/2018 | WO | 00 |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/US2017/033882 | May 2017 | US |
| Child | 16615788 | US |
