Research progress on neutralizing epitopes and antibodies for the Rabies virus

Abstract

Rabies is a zoonotic infectious disease with a high fatality rate. It is caused by a virus in the genus Lyssavirus and is a global public health threat. The rabies virus invades and infects cells mainly via a glycoprotein, which may involve multiple receptors. Neutralizing antibodies against the rabies virus function by blocking the binding of the glycoprotein to a receptor or preventing the membrane fusion process. Vaccination combined with anti-rabies virus neutralizing antibodies is essential for postexposure prophylaxis for category III exposure to the rabies virus. In this review, we discussed the neutralizing epitopes of the rabies virus and the neutralization mechanism of monoclonal antibodies. The neutralizing antibodies that have been commercialized or are under development are also summarized. Our review would provide a basis for the further development of safe and effective broad-spectrum neutralizing antibodies to replace the rabies virus immunoglobulin in rabies post-exposure prophylaxis.

Rabies is a zoonotic disease with a fatality rate of almost 100%. Approximately 59,000 people die of rabies every year worldwide, with the vast majority of deaths occurring in the developing countries of Asia and Africa, and 50% of these deaths were children under 15 years of age [1]. In some economically developed countries, the occurrence of rabies has been effectively controlled through the vaccination of dogs. In less developed areas, the prevention and control of rabies are hampered because of the low levels of medical treatment available, insufficient reserves of vaccines and immunoglobulins, and the lack of attention from the public, resulting in a large number of reported cases of rabies [2,3]. The causative agent of rabies, the rabies virus (RABV), is highly neurotropic and all mammals are susceptible to the virus. The transmission of RABV is mainly through bites or scratches from infected animals, usually dogs. In recent years, the RABV has also been transmitted by bats, foxes, and other wild animals in some countries, such as the United States and Russia [4,5]. RABV-related lyssaviruses can also cause rabies and have a similar mortality rate to RABV. Sporadic cases of rabies in humans caused by EBLV1, EBLV2, ABLV, IRKV, DUVV, and MOKV have been reported in recent years [6], [7], [8].

Rabies is preventable via vaccination and treatable in the early stages after infection. Rabies exposure can be classified into three categories based on the exposure extent. Category Ⅲ exposed persons, those who are bitten or have damaged skin contaminated by the saliva of potentially infected animals, require immediate wound washing plus the administration of a rabies post-exposure prophylaxis (PEP) regimen consisting of vaccination and rabies immunoglobulin (RIG) according to the guidelines of the World Health Organization (WHO) [1]. The timely administration of PEP has been proven to be highly effective against RABV infections. Cases have been reported of illness and death in category Ⅲ exposed persons who received only the rabies vaccine [2,9,10]. Worldwide, 19–50 million people receive PEP every year, resulting in $1.5 billion in economic burden [1]. Currently, human polyclonal anti-rabies immune immunoglobulins (HRIGs) or equine polyclonal anti-rabies immunoglobulins (ERIGs) are used in the PEP regimen, the cost of which is high because of the limited supply, high manufacturing costs, and relatively short shelf life of HRIG and ERIG, making PEP unaffordable in many impoverished areas. Other problems arising from the production of RIG include batch-to-batch variations and the potential risk of bloodborne pathogens or agents [9]. Additionally, ERIG may cause severe allergic reactions. In some developing countries, only approximately 2% of patients with category Ⅲ exposure could receive RIGs [11].

The development of safe, low-cost, effective, and cross-protective RIG alternatives is desirable to stop rabies in humans, particularly in developing areas. This paper reviews the epitopes of the RABV, progress in the development of neutralizing antibodies, and discusses the neutralization mechanism, providing important information for the development of efficacious neutralizing antibodies as RIG alternatives with broad-spectrum activity and a low cost.

1. Etiology of the RABV

RABV belongs to the Rhabdovirus family, Lyssavirus genus, and is the prototypic lyssavirus. The lyssaviruses are classified into phylogroups based on genetic and antigenic data (Table 1) [12]. All the members of the Lyssavirus genus are highly neurotropic, and can cause fatal encephalitis, similar to rabies [13]. Theoretically, all mammalian species are susceptible to lyssavirus infections [14]. Infection of humans with lyssaviruses other than RABV is very rare, with only 13 documented human fatal cases [15]. Most infections caused by lyssaviruses other than RABV have been found in dogs, cats, civets, bats, and other animals [16], [17], [18], [19].

Table 1.

Classification of the Lyssavirus genus.

Species	Abbreviation	Serotype	Reservoir host
Phylogroup Ⅰ
Rabies Lyssavirus	RABV	Serotype 1	All mammals
Australian Bat Lyssavirus	ABLV	Serotype 7	Bats
Duvenhage Lyssavirus	DUVV	Serotype 4	Bats
Aravan Lyssavirus	ARAV	–	Bats
Bokeloh Bat Lyssavirus	BBLV	–	Bats
Irkut Lyssavirus	IRKV	–	Bats
Khujand bat Lyssavirus	KHUV	–	Bats
Gannoruwa Bat Lyssavirus	GBLV	–	Bats
European Bat 1 Lyssavirus	EBLV-1	Serotype 5	Bats
European Bat 2 Lyssavirus	EBLV-2	Serotype 6	Bats
Taiwanese Bat Lyssavirus	TWBLV	–	Bats
Kotalahti Bat virus	KBLV	–	Bats
Phylogroup Ⅱ
Shimoni Bat Lyssavirus	SHIBV	–	Bats
Lagos Bat Lyssavirus	LBV	Serotype 2	Bats
Mokola Bat Lyssavirus	MOKV	Serotype 3	Rodents and domestic animals
Phylogroup Ⅲ
West Caucasian Bat Lyssavirus	WCBV	–	Bats
Ikoma Lyssavirus	IKOV	–	African civet
Lleida Bat Lyssavirus	LLEBV	–	Bats

The current vaccines and biologics have been developed mainly against RABV and the protective effect against non-RABV lyssaviruses is correlated with the antigen distance between the virus strain and the vaccine strain. Current licensed vaccines, such as human diploid cell vaccine and Vero cell purified rabies vaccine, have been shown to provide partial or complete protection against phylogroup Ⅰ lyssaviruses, depending on the challenge pathway, and mild efficacy against phylogroup Ⅱ and phylogroup Ⅲ lyssaviruses [20], [21], [22], [23].

The RABV contains a single-stranded, negative-sense RNA genome with a length of approximately 12 kb, and 5 structural genes are sequentially arranged from the 3′ end to the 5′ end, encoding nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), and RNA-dependent RNA polymerase (L). Nucleoproteins bind to genomic RNA to form a template for transcription and replication and protect genomic RNA from the innate immune response and intracellular nucleases. N, P, and L proteins combine to form a highly helical ribonucleoprotein complex, which is encapsulated by a viral capsid consisting of M and G proteins to form intact virions. The glycoprotein of RABV (RABV G) mediates membrane fusion and plays a pivotal role during the viral infection process [24], [25], [26].

RABV G is composed of 524 amino acids with a molecular weight of 65 kD and an isoelectric point of 6.70. The signal peptide composed of the first 19 amino acids is excised in Golgi bodies, and then the protein is glycosylated in the endoplasmic reticulum to form the mature RABV G with 505 amino acids and consisting of three parts: the ectodomain; the transmembrane domain; and the cytoplasmic domain. The ectodomain, aa 1–439, is the largest domain and forms homotrimers, known as the "spikes", and is the authentic binding region for receptors. Yang et al. were the first to solve the crystal structure of the G-ectodomain, which can be divided into three domains: the fusion domain (FD), pleckstrin homology domain (PHD), and central domain (CD). In this structure, the CD folds back into the PHD/FD junction region, resulting in a bent hairpin conformation of approximately 80 Å at basic pH. Studies on a vesiculovirus glycoprotein have described four conformational states: pre-fusion; early intermediate; late intermediate; and post-fusion states. The membrane fusion process occurs in acidic environments, and a change in pH from basic to acidic results in extensive reorientation of the 3 domains from the bent hairpin conformation to an elongated conformation that shifts the RABV G from a pre-fusion state to a post-fusion state [27,28]. The ectodomain can bind to a variety of receptors and mediate viral infection and migration in different locations or tissues of a host [29]. The transmembrane domain located at aa 440–461 is associated with viral lipid membrane fixation of the RABV G. The cytoplasmic domain located at aa 462–505 interacts with the M to complete the assembly of the virus [25,26,30,31].

2. Neutralizing epitopes of the RABV

Studies have demonstrated that the RABV N has group-specific epitopes shared by all the RABVs, and comprises 4 antigenic sites. Antigenic site I (aa 358–367) and IV (2 independent separate regions, aa 359–366 and aa 375–383) on the N protein are composed of linear epitopes, whereas antigenic sites II and III are conformational epitopes without precise locations [32]. Several epitopes of sites I and IV composed of aa 358–367 are highly conserved across RABV and RABV-related lyssaviruses. Studies have shown that antibodies targeting N protein epitopes do not have neutralizing effects. Therefore, monoclonal antibodies (mAbs) against N protein epitopes have been used for the diagnosis and discrimination of RABV and RABV-related lyssaviruses around the world [33,34].

Liu et al. have identified a variable epitope (PPYDDD) in the M protein, which is located at aa 25–30, and anti-M protein mAbs can be used to discriminate between different RABV strains [35]. Zhao et al. have prepared three mAbs (L-C, 3F3, and 3A6) against l-protein and identified three epitopes. The mAb l-C recognized the conservative epitope 1724VFNSL1728, and the mAbs 3F3 and 3A6 recognized 2 variable epitopes 1479EIFSIP1484 and 1659RALSK1663, respectively. These 3 mAbs respond differently to RABV variant strains and can be used to define RABV strains [36]. The P protein epitope VLGWV is located at aa 262–266 in the C-terminal domain. W265 in this epitope is located in the plane of this domain and may be the key amino acid for the functioning of the P protein. The C-terminus of the P protein binds to the N-RNA template and is important for viral transcription and replication. The intracellular antibody RV1A2 against the C-terminal domain of the P protein can effectively inhibit the replication of RABV in MNA (mouse neuroblastoma) cells. However, it is unclear how antibodies against P protein epitopes play a role in inhibiting viral replication, although the antibodies may perhaps interfere with the P protein binding to the nucleoprotein-RNA template. Antibodies against P protein epitopes may be useful in the development of antiviral drugs [37], [38], [39].

The RABV G is the sole target for all currently known neutralizing antibodies, and therefore, the antigenic structure of the RABV G has been extensively studied and four major antigenic sites (Ⅰ, Ⅱ, Ⅲ, and Ⅳ) and a minor antigenic site “a” have been determined as depicted in Figure 1. Antigenic site Ⅰ is a highly conserved linear epitope composed of aa 226–231. Marissen et al. have demonstrated that the mAb CR57 had reactivity with three polypeptides KLKLCGVL, LKLCGVLG, and KLCGVLGL, and indicated that there was a minimal binding region comprising 6 residues (KLCGVL, aa 226–231) in RABV G, with K and CGV being the key residues [40]. However, the RABV G antigenic sites of some escape mutants often show amino acid substitutions, which not only prevents an antibody binding to the epitope with the amino acid replacement, but also may affect other antigenic binding sites for antibodies because of conformational changes in the RABV G. An ERA (Evelyn-Rokitnicki-Abelseth) escape variant screened by the mAb M725 showed a point mutation (L231P), and 2 RABV strains isolated from bats, BBBV and SHBV, containing P231 failed to bind to M725, demonstrating the key role of leucine at residue 231 [41].

Fig. 1.

Schematic diagram of neutralizing epitopes on RABV G. The upper panel shows the location of each epitope in the RABV G sequence. Antigenic sites Ⅰ, Ⅱ, Ⅲ, Ⅳ, and “a” are shown in yellow, purple, red, blue, and cyan, respectively. S-S indicates a disulfide bond. Antigenic site II is a discontinuous conformational epitope linked by the disulfide bond. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

The antigenic site Ⅱ contains discontinuous conformational epitopes consisting of 2 distinct regions (aa 34–42 and aa 198–200) linked through a disulfide bond [42], which is highly conserved across RABV strains and is located in the PHD of RABV G. Sun et al. have constructed a Fab antibody library for RABV G, and the antibodies obtained by screening were all directed toward the antigenic site Ⅱ of RABV G [43]. Aavula et al. have constructed a phage display library of single-chain antibody fragments (scFv), and further selected and expressed the recombinant scFv fragment A11 with the strongest binding activity, which was shown to recognize epitopes of the RABV G protein within antigenic site Ⅱ [44].

Antigenic site Ⅲ is a conformational epitope composed of aa 330–338, with aa 333–338 being the main binding region for neutralizing antibodies, among which aa 330, 333, 336, and 338 are critical residues [45,46]. Studies have demonstrated that a mutant RABV strain with the amino acid substitutions D330K and M333R lost its pathogenicity toward adult mice [47]. The escape variants generated with the mAbs 16AH8 and M785 showed a point mutation from lysine to asparagine at position 336. These findings indicated the functional significance of these residues for neutralizing antibody binding to RABV G [41].

Antigenic site Ⅳ, also known as the G5 epitope, was found to harbor a linear epitope located in the CD of the RABV G. A synthetic peptide containing this region induced broadly neutralizing antibodies after animal immunization and bound to the monoclonal antibodies 2–22C5 and 615C4 [48], [49], [50]. Cai et al. have found that the minimal binding region of the G5 epitope was located at aa 261–264, of which the residues HDF (aa 261–263) were identified as key residues by alanine substitution scanning. The mAb AR16 recognized the triplet of central residues (H261, D262, and F263), and the residue R264 was involved in the binding of AR16 to the epitope. The critical residues HDF are highly conserved in phylogroup Ⅰ but not in phylogroup Ⅱ [51]. Sequence analysis of an escape mutant showed that a point mutation from arginine to histidine at position 264 of the G5 site produced neutralizing resistance [49].

Amino acid residues at 342–343 constitute the minor neutralization site “a”, which is an independent epitope despite its proximity to antigenic site Ⅲ [52].

In addition to the major antigenic sites described above, some isolated epitopes have also been reported. Luo et al. have found that tryptophan at position 251 was critical for the binding of mAb15–13 to RABV G [53]. Yang et al. have used dog serum to screen the epitopes in a phage display peptide library, and found that the segment of aa 126–141 containing the RYDD-W-T motif might be a key epitope region [54]. Furthermore, aa 14–19 at the N-terminus of the RABV G may be a linear epitope and is conserved in phylogroup Ⅰ [55]. In addition, Buthelezi et al. have found several cross-neutralizing epitopes (GUVTTTF, WLRTV, REECLD, and EHLVVEEL), which may contribute to the development of cross-neutralizing broad-spectrum mAbs [56].

The crystal structure of the RVBV G ectodomain has been resolved and the structure of known antigenic sites can be determined. In the RABV G structure (Fig. 2), the antigenic sites Ⅰ, Ⅱ, and Ⅲ are not in close proximity, but previous studies have shown that mAbs targeting the antigenic sites Ⅱ and Ⅲ compete for binding to the RABV-G [57]. Antigenic site Ⅰ was structurally the most prominent epitope of RABV G in the prefusion state and is highly conserved. MAbs targeting antigenic site Ⅰ have been widely used in preparations to develop passive immunity [56]. Evans et al. have shown that antigenic sites Ⅰ, Ⅱ, and Ⅲ may play crucial roles in neutralization and phylogroup I may be more dependent on antibodies targeting antigenic sites I and Ⅲ, while the neutralization of phylogroup Ⅱ is more dependent on neutralizing antibodies targeting antigenic site Ⅱ [12]. Identification of conserved regions and possible cross-neutralizing epitopes of the RABV G are essential for the development of mAbs with extensive coverage across all RABV strains and even all lyssavirus types.

Fig. 2.

Structure of known neutralization epitopes in the RABV G monomer. The ectodomain of RABV G is divided into three domains: fusion domain (FD); pleckstrin homology domain (PHD); and central domain (CD), shown in cartoon. Antigenic sites Ⅰ and Ⅱ, located in the PHD, are shown as yellow and purple spheres, respectively. Antigenic sites Ⅲ, Ⅳ, and “a” are located in the CD and are shown as red, blue, and cyan spheres, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3. Neutralization mechanism of anti-RABV antibodies

All lyssaviruses infect cells through envelope glycoproteins. First, the glycoprotein binds to the cellular receptor to enter target cells to initiate the virus infection process. The acidic environment of the endosome induces a conformational change in the glycoprotein from the bent hairpin conformation to an elongated structure, resulting in the fusion of the viral envelope and endosomal membrane, and then the nucleocapsid is released into the cytoplasm to complete the infection process [58]. Neutralizing monoclonal antibodies (nAbs) are key therapeutic agents for rabies PEP. Anti-RABV human nAbs can be obtained from antigen-specific B cells from the peripheral blood mononuclear cells of vaccinated individuals. NAbs can also be obtained by screening naïve or synthetic phage-displayed antibody libraries. All nAbs target the G protein, including G protein subdomains, but the mechanisms underlying the inhibition of infection are divergent. Neutralizing antibodies are now widely believed to work in several ways. First, by directly blocking the virus from binding to its receptor and second, by preventing the fusion of the viral envelope and the endosomal membrane. In addition, neutralizing antibodies may inhibit the release of virus particles from the surface of infected cells. Neutralizing antibodies can also induce the lysis of infected cells by complement-dependent cytotoxicity (CDC) and can also play a role in antibody-dependent cellular cytotoxicity (ADCC) and opsonization [59].

Blocking the binding of the viral glycoprotein to the cell receptor is the most common neutralization mechanism. However, the absence of a crystal structure for the RABV G in complex with a cellular receptor has hindered the identification of the critical receptor for mediating RABV infections. Several proteins have been proposed to act as RABV receptors.

The nicotinic acetylcholine receptor (nAChR) was the first protein suggested to be a RABV receptor. Studies have shown that the nAChR binding site of RABV G was located at aa 173–204, corresponding to antigenic site Ⅱ [60]. Thus, neutralizing antibodies targeting antigenic site Ⅱ may block the binding of nAChR and RABV G, preventing the cell from being infected by the virus. However, there is no direct evidence from animal models that indicates that the nAChR acts as a RABV receptor. Furthermore, RABV can infect neurons that do not express nAChR, and nAChR is located mainly on the muscle side of neuromuscular junctions [61]. Thoulouze et al. have found that there was a high expression of the nerve cell adhesion molecule (NCAM) on the surface of RABV susceptible cells, and the expression level was decreased after virus infection. When the NCAM soluble protein, virus, and cells were co-incubated, the virus infection was inhibited [62]. However, the CVS (challenge virus standard) strain can infect Huh-7 cells and A549 cells that do not express NCAM [63]. These studies suggested that the NCAM receptor is not a RABV-specific receptor. In addition, the nerve growth factor receptor (p75NTR) has been found to interact directly with the ectodomain of trimerized RABV G [64,65]. Gluska et al. have reported that RABV transport that was dependent on p75NTR was faster and more directional than transport that was independent of p75NTR [66]. The binding of RABV and p75NTR required K330 and R333 of the RABV G, which corresponds to antigenic site Ⅲ [67]. Yang et al. have analyzed the crystal structure of mAb 523–11 in complex with RABV G, and found that mAb 523–11 recognized antigenic site Ⅲ of the RABV G, and the residue R333 within site Ⅲ played an important role in the recognition between p75NTR and antigenic site Ⅲ [28]. Therefore, the binding of mAb 523–11 to RABV G may hinder the binding between p75NTR and RABV G. However, knockout mice deficient in the p75NTR extracellular region were still susceptible to RABV, and only 25% of primary dorsal root ganglion cells expressing p75NTR were infected, suggesting that p75NTR may only play an assisting role in RABV G binding to the target cell membrane [68]. Recently, metabolic glutamate receptor subtype 2 (mGluR2) has been identified as critical for RABV infection. Wang et al. have found that mGluR2 interacted with RABV G to mediate virus invasion and infection, and RABV together with mGluR2 were internalized into target cells and transported to early and late endosomes [69]. However, African green monkey kidney cells (BS-C-1), which do not express any of the above 4 factors, were still susceptible to RABV infection [70]. N-methyl-d-aspartic acid, glutamate, and dopamine receptors may also be RABV receptors [71,72]. Although several proteins have been proposed to act as RABV receptors, the definitive RABV receptors determined from direct evidence remain to be identified. Studies on RABV receptors show that the neurotropism of RABV involves multiple receptors, and there may be synergistic effects between different receptors. RABV may be concentrated by the nAChR at the neuromuscular junction, while the NCAM receptor may concentrate RABV at some parts of the neuronal membrane, allowing RABV G to bind to other molecules on the membrane surface [73]. Then, RABV may combine with the p75NTR in the form of intact virions for retrograde transport along the axon.

In addition to blocking the binding of the virus to the receptor, neutralizing antibodies can interfere with the membrane fusion process. Membrane fusion requires specific conformational changes of the RABV G. When exposed to an acidic environment, the RABV G undergoes a considerable reorientation of the 3 domains and the conformation is altered from the curved hairpin conformation to the extended conformation. Hellert et al. have found that the mAb RVC20 could lock the RABV G in the pre-fusion state or early intermediate conformation, preventing membrane fusion from occurring, which is an essential process in the life cycle of the RABV [74]. Furthermore, Kim et al. have demonstrated that 2 broad-spectrum mAbs—NP-19–9 and 11B6—could inhibit syncytia formation in a CHO cell line expressing RABV G, suggesting that viral genome entry into the cytoplasm could be inhibited by blocking membrane fusion [75]. Because the neurotropism of RABV involves multiple receptors, blocking membrane fusion using neutralizing antibodies may be more efficacious than blocking receptor binding. Therefore, in the selection of candidate antibodies for PEP, mAbs that block membrane fusion may be able to minimize the effect of escape mutant strains and achieve extensive protection against multiple RABV and RABV-related lyssaviruses.

In addition to blocking virus entry into cells, antibodies may interfere with the release of virions or fight viral infection via ADCC or CDC. However, to date, there have been no reported anti-RABV nAbs that inhibit the release of progeny virions. Furthermore, there is no direct evidence either in vitro or in animal models that nAbs resist RABV infection through ADCC or CDC.

4. Progress in the development of neutralizing antibodies against rabies

The high cost and potential risks associated with RIG have hindered the timely use of PEP in category Ⅲ exposed patients in developing countries, resulting in a large number of reported human rabies cases. Extensive studies have been conducted to investigate effective alternatives for RIG. In recent years, safe, low-cost, and high-affinity mAbs have been developed, some of which have progressed into clinical trials or been commercialized (Table 2). SII RMAb is a recombinant human anti-RABV mAb developed by the Serum Institute of India, which targets a highly conserved epitope of the RABV G. Gogtay et al. have conducted a phase 2/3, randomized, single-blind trial in patients with category Ⅲ exposure from suspected rabid animals to evaluate the anti-rabies potential of SII RMAb. The results indicated that a PEP regimen with the SII RMAb and rabies vaccine was safe and not inferior to that using HRIG and rabies vaccine; thus, using the SII RMAb may be a safe and effective alternative to HRIG in PEP [11]. The SII RMAb was approved for marketing in India in December 2016. In addition, ormutivimab injection, consisting of the recombinant humanized mAb NM57, which targets antigenic site Ⅰ, has been developed by North China Pharmaceutical Company and was approved in China on January 25, 2022, which made ormutivimab the first anti-RABV injection to be approved in China. Zhai et al. have shown that ormutivimab could neutralize 18 Chinese prevalent RABV strains in a FAVN (fluorescent antibody virus neutralization) test and had a broad-spectrum neutralization capacity. In addition, ormutivimab showed the same protective effect as HRIG in three challenge models, BALB/c mice, KM (Kunming) mice, and beagle [76]. In phase Ⅲ clinical trials, ormutivimab combined with a human rabies vaccine achieved the primary and secondary efficacy endpoints for category Ⅲ exposed persons, demonstrating that ormutivimab may be an effective and safe RIG substitute. However, mAbs that target distinct epitopes of the RABV G differ in the neutralizing activity across RABV strains and may lose neutralizing ability with the development of escape variants. The WHO Collaborating Centers for Rabies Research have proposed a rabies PEP regime that uses a cocktail of different mAbs that target distinct epitopes without competing for RABV G-binding [77]. Such a mAb cocktail needs to be cross-reactive against all RABV strains, and should even have cross-immunity against RABV-related lyssaviruses because emerging lyssaviruses are being frequently reported worldwide [56,78,79].

Table 2.

Monoclonal antibody injections currently under development or on the market.

Monoclonal antibody injection	Company	Neutralization antigenic site	Current progress
SII RMAb	Serum Institute of India	Ⅱ	Commercialized in India
Ormutivimab Injection	North China Pharmaceutical company	Ⅰ	Approved in China
Twinrab™	Zydus Cadila of India	Ⅱ/Ⅲ	Commercialized in India
NM57S/NC08	North China Pharmaceutical company	Ⅰ	Ⅰ/Ⅱ clinical trial
SYN023	Synermore Biologics	Ⅲ	Ⅲ clinical trial
CL-184	Crucell Holland BV	Ⅰ/Ⅲ	suspend
RVC20 and RVC58	Institut Pasteur	Ⅰ/Ⅲ	Preclinical study
NP-19–9 and 11B6	Celltrion, INC	Ⅱ/Ⅲ	Preclinical study

Extensive studies on mAb cocktail combinations targeting multiple antigenic sites have been performed, and these combinations have shown satisfactory results in vivo, in vitro, and in clinical trials [[80], [81], [82], [83]]. Muller et al. have developed a mAb cocktail consisting of five murine mAbs, of which E559.9.14, 1112–1, M727–5–1, and M777–16–3 recognize antigenic site Ⅱ of RABV G and 62–71–3 recognizes antigenic site Ⅲ. Evaluation of the potency of this mAb cocktail in vivo and in vitro, indicated that it produced a protective effect comparable with that of HRIG [84]. Twinrab, a murine monoclonal antibody mixture developed by ZydusCadila of India, contains two monoclonal antibodies 62–73–1 and M777–16–3, which target epitopes within antigenic sites Ⅱ and Ⅲ, respectively. Kansagra et al. have conducted a phase 3, randomized, open-label, non-inferiority trial using Twinrab in category Ⅲ suspected rabies exposed persons. The results showed that the geometric mean RFFIT (rapid fluorescent focus inhibition test) titers of the Twinrab arm and the HRIG arm on day 14 were 4.38 and 4.85 IU/mL, respectively, demonstrating that Twinrab was not inferior to HRIG [85]. Twinrab was approved in India in September 2019. Although murine-derived mAbs can specifically recognize antigens and have the advantages of low cost and a relatively high safety profile, these mAbs cannot effectively activate effects related to the complement and Fc receptors in humans, and the potential human anti-murine antibody immune response restricts their clinical application.

With developments in genetic engineering and single cell technology, most monoclonal antibodies intended for clinical development are humanized. CL-184 is a 1:1 cocktail of the mAbs CR57 and CR4098, which target antigenic sites Ⅰ and Ⅲ, respectively, and were produced using the PER. C6 cell line. Goudsmit et al. have demonstrated that CL-184 was able to neutralize all 26 RABV isolates tested in vitro and also provided a similar protection rate to HRIG against a lethal RVBV challenge in Syrian hamsters [86]. Bakker et al. have conducted a clinical trial of CL-184, which showed that the RVNA (rabies virus neutralizing antibody) levels of all subjects were higher than 0.5 IU/mL on day 14 after administration of CL-184 combined with a rabies vaccine [87,88]. In addition, there were no escape mutant strains observed in vitro under the selection pressure of the mAbs CR57 and CR4098, and no natural isolates showed mutations in either epitope recognized by CR57 and CR4098, demonstrating the neutralization potential of the cocktail CL-184 [89]. However, the mAb cocktail CL184 was less effective at neutralizing RABV-related lyssaviruses, compared with the commercial HRIG.

Emerging RABV-related lyssavirus infections with symptoms similar to RABV infections are being reported worldwide, and therefore it is important to expand the protective spectrum of current anti-RABV cocktails. SYN023, an antibody cocktail developed by Syndrome Biologics, consists of the 2 humanized mAbs CTB011 and CTB012. When these mAbs were mixed in a ratio of 1:1, SYN023 showed the same or better neutralizing activity, compared with HRIG, against 10 North American wildtype RABV isolates and 15 Chinese RABV strains in vitro. At a dose of 0.03 mg/kg, SYN023 provided the same level of protection as standard HRIG administration (20 IU/kg) in Syrian hamsters infected with highly virulent bat RABV variants. The results of sequence analysis indicated that CTB012 is expected to be able to neutralize many RABV-related lyssaviruses [80]. SYN023 is currently undergoing phase Ⅲ clinical trials and is expected to become the world's first commercially approved humanized antibody cocktail against RABV. Kim et al. have developed an antibody cocktail composed of the antibodies NP-19–9 and 11B6, which target antigenic sites Ⅱ and Ⅲ respectively. This cocktail was able to neutralize 47 lyssaviruses in vitro, including EBLV-1, EBLV-2, ABLV, and DUVV (P. K. [75]).

Although the antibodies described above are effective in rabies prevention, no studies have investigated whether they can be used in the clinical treatment of rabies. The mAbs RVC20 and RVC58 targeting antigenic sites Ⅰ and Ⅲ, respectively, were screened and expressed in 293 Freestyle cells. In vitro, a cocktail containing RVC20 and RVC58 was able to neutralize all 35 pseudotyped RABV strains, and 25 RABV-related pseudotyped lyssaviruses, indicating excellent potency and breadth of reactivity [90]. Furthermore, de Melo et al. have reported that the concomitant administration of RVC20 and RVC58 mAbs could clear the virus from the central nervous system, normalize the pro-inflammatory cytokine expression in the brain, and abolish symptoms in rabid mice through simultaneous intracerebroventricular infusion and intramuscular injection at the infected site, indicating the potential for this mAb cocktail to be an effective treatment for symptomatic rabies cases [91].

Nanobodies, also called VHH antibodies, are antibodies containing only the smallest antigen-binding fragments of the heavy chain, which have the advantages of both conventional antibodies and small molecular antibodies, including low production costs and good thermal stability; therefore, nanobodies may be ideal antibody drugs. Nanobodies targeting different epitopes can be fused to generate multivalent molecules to improve the neutralizing ability and prevent mutant virus strains from escaping. Nanobodies with an extended half-life against RABV have been reported to be feasible for PEP in mouse models [79,92,93]. In addition, the nanobodies have a small molecular weight and can penetrate the blood-brain barrier, which may lead to the development of therapeutic antibody drugs for rabies treatment.

5. Summary and prospects

All 5 structural proteins of the RABV can induce the production of monoclonal antibodies, but the antibodies targeting epitopes of proteins other than RABV G are mainly used for the diagnosis of RABV and RABV-related lyssaviruses. RABV G, which is located on the envelope surface of the virus and containing the antigenic sites Ⅰ, Ⅱ, Ⅲ, Ⅳ, and “a”, is the only target to induce neutralizing mAb production. The RABV G can combine with a variety of receptors, which are directly related to virus infection and invasion. To date, a variety of mAbs targeting RABV G epitopes have been developed, and multicomponent mAb cocktails have displayed better potency than HRIG in vitro, in vivo, and in clinical trials. However, some key residues in antigenic sites may undergo amino acid substitutions, leading to the development of neutralization resistance and escape mutants. In addition, many antibodies have displayed limited neutralizing activity against RABV-related lyssaviruses, especially phylogroup Ⅱ and phylogroup Ⅲ lyssaviruses.

Studies have suggested that there may be more antigenic sites in the RABV G than are currently known, which may lead to the screening of more effective neutralizing antibodies as substitutes for RIG in PEP. In addition, although many antibodies have been developed, the mechanism of neutralization remains unclear. Proposed mechanisms for the neutralization include blocking the binding of the RABV G with receptors, and interfering with the membrane fusion process. Specific receptors for RABV have not yet been identified, and the evidence suggests that RABV infection may involve multiple receptors. Antibodies that block the fusion of the viral envelope and endosomal membrane may have a broader neutralizing spectrum and better neutralizing efficacy, compared with antibodies that target receptors. Therefore, it is important to explore the process of RABV infection, and the mechanism of antibody neutralization, and to determine conserved and pathogenicity-related neutralizing antigenic sites. These studies will contribute to the future development of safe, low-cost, effective, and cross-reactive mAb cocktails as an alternative to RIG in PEP, which has important implications for the prevention and control of rabies.

Funding

This work was supported by the National Key Research and Development Program of China (No.2019YFC1200500), the National Natural Science Foundation of China (No. 31902308, 32170154), the Shandong Provincial Natural Science Foundation (No. ZR2019QC010), the Shandong Key Research and Development Program (No. 2019GNC106006).

Author contributions

SC, SP, and ZW wrote the manuscript. YP prepared the tables. The figures were prepared by SC. LL, TL, LW, and WM compiled the sections. ZX developed the outline of the report. All authors contributed to the final manuscript.

Acknowledgement

None.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data available statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Ethics statement

Ethics approval were waived for this study because no patients’ data were reported.

Informed consent

Informed consent was waived for this study because no patients’ data were reported.