Structural Basis of Enveloped Zika Virus Specific Antibody Protection

Abstract

ZIKV as an enveloped, positive-sense, single-stranded RNA virus, belonging to the family of flavivirde.(1-3) The rapid geographical expansion and the rampant effect of zika virus (ZIKV) in the recent years have posed serious threat to human life, which lead World Health Organization to declare a Public Health Emergency of International Concern (PHEIC) against this flavivirus in 2016.(4) Binding and fusion of a virus with the host membrane is the first and one of the most crucial steps in the life cycle of all enveloped viruses. Therefore, it is essential to understand the structure and pathogenicity in order to develop therapies for prevention. The cryo-EM structure of ZIKV helped elucidated the structural proteins of ZKIV. The structural protein is mainly composed of pre-membrane (prM/M), envelope (E) and capsid (C) protein. All these proteins are important components of fusion and binding events into host cells. Current research indicates a fusion event takes place in the acidic endosomal environment which triggers an irreversible conformational change of ZIKV’s three-domain envelope glycoprotein (E).(6) With the help of structural infection of ZIKV, researchers are trying to develop neutralizing antibodies against ZKIV to abate the infection and reduce the risks from ZIKV. The envelope protein is of particular interest because it helps mediating viral assembly, entry, and fusion hence being the main antigenic target for neutralizing antibodies. Therefore, development of neutralizing antibodies for the E protein could be the best therapeutic approach to offer protection against infection.

Introduction

Zika virus (ZIKV) is a mosquito borne virus first isolated in 1947 from a sentinel rhesus monkey in Uganda. This single stranded RNA virus has been associated to many other viruses like dengue virus, Chikungunya virus and West Nile virus, as they belong to the same the genus of Flavivirus.(9) The main transmission source of ZIKV is Aedes aegypti mosquito but some additional means of transmission are sexual, congenital or in some cases by blood transfusion. In usual cases 80% patients are asymptomatic for 3-12 post exposing to infection. The most commonly reported symptoms are fever, rash, headache and fatigue. (10-11) ZIKV is also linked to some severe symptoms where neonates were reported with microcephaly and increase in congenital CNS malformations. Furthermore, reports have suggested adults suffering from Gullain-Barre syndrome and meningitis. This indicates a strong link between ZIKV infection and CNS abnormalities. Recently, there has been upsurge in ZIKV infection and became the matter of concern do to the harm ZIKV can pose to human life.

Although its first major outbreak was in 2007 on Yap island and Micronesia, the recent emergence of ZIKV in 2017 raised global concerns, by affecting millions in 84 different countries including the United States. (14) Furthermore, the World health organization (WHO) declared a public health emergency of international concern in 2016, when ZIKV infection was associated with microcephaly and Guillain barre syndrome. (15) Knowing the potential threat of ZIKV to the mankind, researchers are study the structure of the Zika envelope protein and develop better vaccines for treating Zika virus.

Like all the flaviviruses, the ZIKV genome is translated into extended polyprotein in the cytoplasm of the host cells. Further this polyprotein is cleaved by viral/host proteases into structural and non-structural proteins. The structural proteins comprise of three major components namely: capsid (C), membrane precursor/membrane (prM/M) and envelope (E) protein which help in forming the virus particles. Further, there are total seven non-structural proteins (NS1, NS2a, NS2b, NS3, NS4a, NS4b, and NS5) responsible for genome replication, handling polyprotein and antagonizing host response.

Along with this, there is an untranslated region bordering in 5’ and 3’ flanking at the end. Additionally, similar to other enveloped viruses, in the life cycle of ZIKV, binding and fusion is one of the key events allowing the entry of virus into the host cell. It is traced that the E protein plays a major role in the event of binding and fusion of the virus. In order to develop vaccines against this virus, it is very crucial to understand the pathogenesis of the virus better. (18) Scientists are trying to elucidate the structure of this virus, and for the first time Richard Kuhn et. al. reported the Cryo-EM structure of mature ZIKV, which offered a lot of information to comprehend the unique features of this virus. It was elucidated, this virus has icosahedral arrangement with of 180 copies of E and M protein each, where E dimers surrounds the capsid. The prM/M has a M-loop and stem & transmembrane region (M-TM) which helps anchor the protein to the host lipid membrane. Furthermore, the E protein is mainly accountable for receptor viral assemble, binding and fusion. It mainly comprises of four domains which are: i) stem (E-stem) and transmembrane (E-TM) domain, ii) Domain I (DI), iii) Domain II (DII) and iv) Domain III (DIII) accountable for different functions. The stem transmembrane domain is responsible for membrane support. In the E protein, b-barrel domain (DI) acts as a bridge between DII and DIII (Figure 2).

Further, the fusion loop (FL) is at the edge of DII has hydrophobic sequence, and is known to interacts with the with the host lipid bilayer during the fusion event. Like all other flaviviruses, DIII comprises of the receptor-binding site also carrying an important role during the fusion (Figure 3). (18-19)

The zika virus has three main phases in its life cycle, i) immature (non-infectious), mature (infectious), and fusogenic (binding phase). The virus is in the non-infectious state when it enters host cell via endocytosis. Due to the acidic environment in the endosomes, the E protein goes under an irreversible conformational change and the dimer is transformed into a trimer leading to fusion event. In the ER, this newly formed virus offspring forms immature virions which comprise of 60 E: precursor-membrane (prM) heterodimers. This is followed by the maturation phase in the Trans-Golgi, where the acidic pH stimulates the restructuring of the E-prM heterodimers into 90 E: M homodimers. The immature structure has its cleavage site of prM exposed, where furin can (a host protease) slices prM into pr and M protein, leading to maturation. In the immature virus pr protects the fusion loop but after slicing the pr, fusion loop is exposed allowing the virus for pH-mediated fusion. (19-21)

Understanding the structural features of the virus is very important in the process of therapeutics development. Monoclonal antibodies (mAb) have been a hot subject for drug discoveries, and for years they have been used for variety of diseases like cancer, inflammatory diseases, neuro-degenerative diseases and infectious diseases. Development of mAb against flaviviruses is a very remarkable strategy to fight the infection. The mAb can safeguard against flavivirus infection at numerous steps which can impeding virus attachment to the host cell membrane, disrupt membrane fusion of virus, or it can be by triggering Fc-dependent effector functions (Figure 4). Although the posed advantages, weak/non-neutralizing antibodies at a sub-neutralizing concentrations may lead to antibody-dependent enhancement (ADE) of infection, an incident where the antibody helps facilitate the entry of virus via the Fc receptor (not expressed in all the cells) and increasing the infection rate.(22)

Therefore, for the flaviviruses, E-protein because of its functions remains a common target for generally known neutralizing antibodies. As discussed earlier, the three domains in E-protein are very important for the receptor identification, attachment and fusion events, hence mAb targeting E-protein can help stop infection by acting at the first step. However, there are antibodies against non-E proteins, mostly non-neutralizing, making them less favorable for the therapeutics approach. Hence, neutralizing antibodies with E protein as the antigenic targeting site are being looked upon as a promising solution in the development of better therapeutics interventions for ZIKV. In this study six monoclonal antibodies (mAbs) were developed for ZIKV by immunizing the mouse with live virus and enhancing with either infectious virus or recombinant E proteins. This report talks about the four of the mAbs (ZV-48, ZV-54, ZV-64, and ZV-67) neutralized infection of ZIKV to varying degrees, and also two ZV-2 and ZV-13 inhibited infection poorly against ZIKV.(23)

Results and Discussion

Testing neutralizing activity of ZIKV in vitro:

The monoclonal antibodies developed were tested against ZIKV E (envelope), ZIKV E-FL (fusion loop mutant), ZIKV DIII, WNV E (West Nile virus envelope), and DENV-4 E (Dengue virus envelope) with the help of ELISA. The results indicated all the mAbs were specific to ZIKV DIII, except ZV-13 having cross reactivity against the WNV E. All the rest of mAb possessed minimal cross reactivity against WNV E and DENV4 E strain. To further perform primary analysis and test the efficiency of mAb, focus reduction neutralization tests (FRNT) were carried out. This is a widely used assay to test the neutralizing capability of the mAbs against specific virus. The test was carried out for various African and American ZIKV strains (H/PF/2013, Paraiba, Dakar, MR-766). Six mAb were incubated along with different strains and the result was measured as relative infection %. The results indicated ZV-13 and ZV-2 failed to neutralize against all the ZIKV strains in comparison to the other mAbs. Furthermore, the results also indicated, ZV-48 and ZV-64 also show reduced neutralization effects in some ZIKV strains. The reason for this is not clear.

Binding affinity of ZIKV mAbs:

So far, the neutralizing efficiency of all the six monoclonal antibodies was tested, but the discrepancies within each mAbs were not addressed. One possible reason for this would be different binding site, so to understand this each mAbs were tested using biolayer interferometry. This is an analytical technique used for measuring molecular interactions. The biosensor tip is immobilized with protein and flowed with the sample. Any interactions would result in wavelength shift due to increase in optical density. This technique helps determine the binding affinities, binding, rate of dissociation and association.

To perform this assay, six mAbs were immobilized on different sensor chip and recombinant DIII was administered in the system. The antibody having maximum interaction with the DIII would indicate they are more neutralizing and further having greater binding affinities for the recombinant protein. Out of all the mAbs, KD equilibrium value for ZV-54 and ZV-67 was less than10 nM in contrast to ZV-64 and ZV-48 with the KD equilibrium ~ 35 nM, demonstrating highest binding affinity of ZV-54 and ZV-67 to the DIII. Furthermore, the half-lives of ZV-54 and ZV-67 was 33 and 13.8 min respectively, indicating the slowest dissociation rates when compared to ZV-64 and ZV-48 with half-lives of 1 and 3.2 min. This experiment further helped determine the discrepancies between all the mAbs and indicated the strength of ZV-54 and ZV-67.

Although some of the mAbs showed neutralization effect, it was still not clear whether they are responsible for antibody dependent enhancement or not. For neutralization, antibodies should be in sufficient stoichiometry, or the antibody help facilitate the virus in and increase the infection. Therefore, to test the degree of ADE and antibody concentration dependency an assay was performed on K562 cells expressing Fcγ receptor II for ZIKV and DENV-2. All the mAbs showed variable level of enhancement in infection, but the most neutralizing mAbs ZV-54 and ZV-67 showed ADE only at sub neutralizing concentrations. This means given an appropriate amount of these mAbs, the chances of ADE could be lowered.

Although the binding affinity of all mAbs had been identified, it was unclear how the antibodies reacted on a structural level. To understand the structural basis of the interaction between the antibodies and DIII of ZIKV X-ray crystal structures were generated for all the DIII with four antibody complexes (ZV-2, ZV-48, ZV-64 and ZV-67). The results indicated, ZV-2 and ZV-67 were binding near heavy chain region, whereas ZV-48 and ZV-64 engaged at DIII by the light chain domains. From the results, the binding of ZV-48 and ZV-64 appeared to be very similar and therefore the further docking studies were carried out to determine the binding site of all the mAbs. Docking results indicated ZV-2 and ZV-67 bound to different epitopes and therefore, should not compete with each other during binding. On the contrary, ZV-67 and ZV-54 recognized same site on DIII leading to competitive binding (Figure 10). To experimentally prove this docking results, competitive binding assay were carried out. To perform the competitive binding, once ZV-67 was immobilized, and ZV-2 and ZV-64 when introduced could easily bind to the DIII, but ZV-54 could not bind portraying both the mAbs bind to the same determinants of the DIII. Similarly, when ZV-48 was immobilized, ZV-67 and ZV-2 captured on the DIII but due to competitive binding, ZV-64 was blocked. All the outcomes reinforced the docking analysis of three different epitope binding site (Figure 11).

In Vitro Studies:

After establishing the ground in vitro and doing computational analysis, the results were indicative about the nature of the neutralizing effect of mAbs against the ZIKV infection. Therefore, to further validate the results, the effect of mAbs were tested in vivo on mice deficient in INF signaling. These mice were subjected to either non-binding control antibodies or anti ZIKV mAbs. In this experiment, two of the most neutralizing antibodies were chosen based on all the results obtained before. The results found indicated, increased level of ZIKV RNA in the serum of mice when treated with CHK-166 (control) in comparison to ZV-54 and ZV-67. Also, when the weight was measured for all the mice, control model showed reduced weight when compared to mice treated with ZV-67 and ZV-54. Furthermore, the survival analysis was also constant with the previous results, indicating reduced viremia and high survival rates in presence of the mAbs (ZV-67 and ZV-54). All the results were indicative of the neutralization capacity of the generated mAbs.

Conclusion

ZIKV has raised global concerns in the past, and due to the lack of vaccines available to treat this infection hence, it is very crucial to understand the structure of the ZIKV and try to develop vaccine for this harmful infection. To take a step further in the vaccine development for ZIKA virus, a group of mAbs were developed against ZIKV which could be a potential vaccine. The research focused on developing mAbs and acquire an understanding of epitopes identified by neutralizing antibodies. Mice were inoculated with ZIKV and the specific mAbs generated against ZIKV were analyzed. All the results suggested, four mAbs had binding affinity to ZIKV DIII, and further neutralized infection of ZIKV Asian strain. Out of those four, ZV-54 and ZV-67 proved to be more potent in neutralizing against other ZIKV strains whereas, ZV-48 and ZV-64 has a lower inhibitory action against African ZIKV and American strains. Furthermore, on analyzing the sequence of ZV-48 and ZV-64 the results indicated they are sibling clones’ due to the similarity in the VL region. In contrast, the findings also suggested ZV-54 and ZV-67 are related mAbs due to the similarity in VL and VH sequences. After the successful identification of the DIII binding epitopes of all the mAbs, in vivo studies carried out implied ZV-54 and ZV- 67 act as a defense means against an African ZIKV strain. Overall, this research introduced newly identified monoclonal antibodies for ZIKV DIII, with neutralizing effect, which could be used as in diagnostic assay or offer direction for development of preventive antibodies for pregnant women, or can also serve as a therapeutic antibody.