Introduction

In December 2019, a severe acute respiratory syndrome (SARS)-like disease of unknown etiology emerged in individuals directly exposed to China’s Huanan Wholesale Seafood Market [1, 2]. Subsequently, the disease, caused by a new coronavirus named Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), spread rapidly and became a pandemic known as COVID-19. The World Health Organization announced the COVID-19 outbreak as a Public Health Emergency of International Concern and has been maintaining this status for over 3 years, with over 760 million confirmed cases and over 6.9 million deaths worldwide (data on July 28, 2023) [3]. Currently, the pandemic is under control, mainly due to population immunity, and the disease is fading to an endemic [4, 5]. Since the virus is circulated in the environment and the population and is unlikely to be eradicated, infection remains a public health concern regarding long-term effects on individual health, human and animal reservoir and transmission, immune-escaping mutation of the virus, and morbidity and mortality of the disease in susceptible individuals.

Besides the role of population immunity in COVID-19 protection, therapeutic agents are required, even after the pandemic, to prevent and reduce disease severity, especially in susceptible individuals. Numerous new and repurposed drugs and herbal medicines have been developed and tested for COVID-19 [6,7,8]. Remdesivir, which inhibits viral RNA synthesis, and Paxlovid, the Mpro inhibitor, are FDA-approved drugs to treat COVID-19 [9]. However, drugs are limited in specific populations, such as pregnancy, breastfeeding, and renal impairment [10]. In addition, high viral mutation rates and drug selection pressure might introduce a drug escape mutation. Therefore, alternative treatments for COVID-19 are required.

Passive immunization with convalescent plasma from recovered patients becomes first-line therapy during the pathogenicity of unrevealed disease. Subsequently, monoclonal antibodies against COVID-19 have been developed and used as part of the therapeutic options for COVID-19, especially in susceptible populations [11,12,13]. The FDA approves some of them, and some are in the development pipeline.

Strategies for developing therapeutic antibodies against SAR-CoV-2 are to reduce viral load or replication by targeting virus proteins that function in the viral life cycle, such as attachment or viral replication. Another strategy is to target host molecules to mitigate the host’s hyperimmune response and disease severity. Another factor that promotes antibody efficiency is the design of the antibody format. Details of strategies for engineering therapeutic antibodies, including antibody format, are reviewed in this article.

Antibody Targeting the SARS-CoV-2 Proteins

SARS-CoV-2, a causative agent of COVID-19, is a betacoronavirus subgroup B in the Coronaviridae family. Coronavirus is an enveloped, nonsegmented, positive-sense, single-stranded RNA virus [14]. The viruses in this family, including SARS-CoV-2, have been reported to infect various animal species, including humans [14,15,16]. Along with SARS-CoV and MERS-CoV, SARS-CoV-2 is one of the three human coronaviruses that cause severe acute respiratory syndrome (SARS) [17, 18].

Antibodies against viral infection usually target the proteins essential in the viral replication cycle. Most antibodies against SARS-CoV-2 targeted a spike (S) protein to block the viral attachment or entry into host cells [19]. Besides the spike protein, other proteins that play an essential role in viral replication are also therapeutic targets, such as Nsp3 (papain-like protease), Nsp5 (main protease, Mpro, 3CLpro), Nsp9, and Nsp12 (RNA-dependent RNA polymerase, RdRp) [20].

Spike (S) Protein

The S protein, composed of two subunits: S1 and S2, forms a homotrimer on the virion surface. The S1 subunit of the virus contains the N-terminal domain (NTD) and receptor-binding domain (RBD), which binds to angiotensin-converting enzyme 2 (ACE2) expressing on the surface of various cell types, including alveolar epithelial cells and oral, nasal, and nasopharynx epithelial cells [21, 22]. The RBD conformation of the spike is interchangeable between the upward (open) and down (close) conformations wherein the ACE2 binding site is exposed and hidden, respectively [23, 24]. S1–ACE2 receptor binding induces a spike conformation change, reveals the S2ʹ site on the S2 subunit, which is cleaved by host TMRRSS2, resulting in shedding of the S1 subunit, and subsequently exposes the fusion loop of the S2 subunit to create the fusion pore-mediated viral genome releases into the host’s cytoplasm [25, 26].

Therapeutic conventional (full-length) antibodies against spike protein constitute a significant group of FDA-approved SARS-CoV-2 antibodies [27]. The aim of targeting spike protein is to protect the virus from entering the cell by directly or indirectly blocking the binding of spike protein to the ACE2. NTD and RBD were reported as therapeutic antibody targets [28,29,30,31,32], although the primary focus was on RBD [27, 28, 32]. The mechanism of antibodies to spike protein includes directly interfering with the ACE2 interaction by occlusion of the ACE2 binding site [32,33,34,35] or acting as a receptor mimic to induce premature S2 fusion loop exposure [36, 37, 38••]. Antibodies can indirectly block RBD-ACE2 interactions by a steric hindrance [28, 39]. RBD-antibody binding in the upward or down conformation or NTD can cause conformational trapping, preventing spike conformational change and hindering viral entry [28, 39,40,41,42]. Antibodies also target NTD by interrupting the trimer formation of spike protein [43]. The S2 subunit is highly conserved across different betacoronavirus lineages [44,45,46]. The antibodies to S2 are divided into two classes: the antibody-targeting fusion peptide and the Sʹ cleavage site [45]. Conformational trapping also occurs in the S2 subunit, preventing fusion loop exposure [41]. S2 subunit targeting is limited by the accessibility of antibodies depending on spike dynamics and disclosure of the epitope [45].

The challenge in developing the antibody-targeting spike protein is the high mutation of the spike in the SARS-CoV-2 variants [19]. The emerging SARS-CoV-2 Omicron variant contains > 30 mutations in the spike protein, especially in the RBD [19]. These mutations caused many available therapeutic antibodies obsolete due to loss or massive reduction of protection against new mutated variants [19, 46]. Amino acid substitutions at positions S477N, T478K, F486V, and E484A decrease the activity of the available anti-spike antibody by 272–10,000-fold [24]. The receptor binding motif (RBM) on the RBD is the most efficient antibody target [30, 38••, 47]. However, RBM is a mutation hotspot that causes loss of antibody activity, especially in Omicron variants [48]. Conversely, although the non-RBM part of the RBD is conserved, the antibody’s efficiency is less than that of targeting the RBM, whereas it tolerated viral escape [30].

Some antibodies endured spike mutations and demonstrated cross-variant protection [32, 38••, 49], which applies to therapeutic antibody design. Bebtelovimab (LY-CoV1404 or 1404), which binds to conserved RBD epitopes, demonstrated cross-variant protection, including Omicron B.1.1.529 and BA.2 [32]. However, the protectivity of the Omicron BQ.1 and BQ.1.1 subvariants is diminished [50]. Anti-RBD spike antibody S2H97 interacted with the spike protein from subgenus Sarbecoviruses and demonstrated broad neutralization across the SARS-CoV-2 variants [38••, 49]. Antibodies developed against SARS-CoV or MERS-CoV have been tested for protection against SARS-CoV-2 [39, 41, 51]. Most of the amino acid residues essential for ACE2 binding by SARS-CoV were conserved in SARS-CoV-2 [52]. Sotrovimab (VIR-7831) is derived from memory B cells of SARS-CoV survivor bound and neutralized SARS-CoV-2 variants [51] and has demonstrated efficacy in early waves of Omicron [53, 54]. Other cross-variant protection antibodies were reported in an antibody that shared the 18 binding-epitope residues and electrostatic contacts on the RBD-binding interface with ACE2 [33] and an antibody to S2 of spike that targeted the highly conserved epitope across different betacoronavirus lineages [44]. Combining two neutralizing antibodies (antibody cocktails), for example, tixagevimab and cilgavimab, bamlanivimab and etesevimab, and casirivimab and imdevimab, demonstrated improved activity/efficiency against mutation escape variants [11, 27, 55••]. Therefore, selecting antibodies that bind to the highly conserved epitope(s) or protect across different lineages of coronaviruses, competing with ACE2 with high similarity, and formulating antibody cocktails can develop as strategies to overcome antibody escape mutations of the SARS-CoV-2 spike. If the critical mutated amino acid responsible for therapeutic resistance in the circulated variant [24] is defined, for example, R436X of the Omicron, designing antibodies targeting the mutant will be another option to develop the broadly neutralizing antibody [56]. However, reevaluation of antibody efficacy is required whenever a new variant emerges [11]. Antibody treatment should be considered to introduce antibody-selected mutations, as reported in high-risk patients treated with sotrovimab [57•, 58]. A single-dose sotrovimab treatment induced E340K/A/V/G and/or P337L/R mutations of Omicron variants, reducing susceptibility to sotrovimab [57•].

Nonstructural Proteins (Nsps)

The viral genome contains 13 open reading frames (ORFs). ORF1a and ORF1b are translated into polyprotein precursors, pp1a and pp1ab. The precursor is cleaved by the viral protease, i.e., Nsp3 and Nsp5, resulting in 16 Nsps that function in viral genome replication and modulation of host immune responses [20, 59]. Nsp12 assembles with Nsp7 and Nsp8 to form a holo-RdRp complex, an essential component for viral RNA synthesis [60, 61]. The holo-RdRp complex coordinates with other accessory subunits, known as replication and transcription complexes (RTC), and promotes the fidelity of RNA synthesis [60, 62]. SARS-CoV-2 Nsps shared structural homology or conserved amino acids/motifs with SARS-CoV and/or other betacoronaviruses [20, 63,64,65,66]. This review focuses on Nsps reported as antibody targets: nsp3, 5, 9, and 12; other SARS-CoV-2 proteins as therapeutic targets were reviewed in [20].

The nsp3 of SARS-CoV-2 is a multidomain protein; among them, the PLpro domain contains cysteine proteolytic (PLpro), deubiquitinating, and deISGylating activities [20]. The protease activity of nsp3 cleaves the pp1a polypeptide to separate nsp1, nsp2, and nsp3 [20]. Additionally, nsp3 suppresses the antiviral immune response by deubiquitination and deISGylation of interferon-stimulating gene 15 (ISG15) [64]. The PLpro domain is the main target for antiviral drug development. In nanobody format, antibodies targeting nsp3 demonstrated inhibition of hydrolytic activity to interfere with deubiquitination, deISGylation, and polyprotein cleavage activities in vitro [64, 67]. However, the ability of these nanobodies to inhibit the authentic virus and interfere with viral replication remains to be investigated.

Mpro, a chymotrypsin-like protease, is a unique protein without human homologs; it is critical in viral replication because it cleaves nsp12 from the polyprotein precursor. Its activity requires homodimerization of the proteins [68]. The Mpro consists of three domains. Domain III functions in homodimerization, allowing domains I and II to form a substrate-binding pocket with the embedded catalytic site [69]. Thus, dimerization inhibition interrupts the enzymatic activity of Mpro. There are three transitional stages during dimerization formation: extended monomeric, compact, and dimeric [69]. Antibody fragments targeting Mpro disrupted dimerization by conformationally trapping Mpro in the predimeric stages [70] or interacting with residues responsible for homodimerization [70]. Cell-penetrating antibodies bound and inhibited the catalytic surface of Mpro and demonstrated the cross-variant protective effect against authentic viruses in cell cultures [70].

Nsp9, an accessory protein in the RTC, undergoes dimerization, RNA binding, and protein recruitment for 5ʹ-mRNA capping, which is essential for viral replication [71,72,73]. Nsp9-bound nanobodies have been reported to induce a topological change [71] or stabilize Nsp9 in a tetrameric form [74], which can interfere with viral replication.

Nsp12, a core component of the RTC [20], is crucial for viral replication and is a target of nucleoside analogs already approved for treating COVID-19 [8]. An antibody to the hepatitis C virus (HCV)-RdRp broadly inhibited viral RNA replication in SARS-CoV-2 variants of concern and other RNA viruses [75].

Antibody to Host Molecules

Cytokine Storms

Inducing uncontrolled inflammation, known as cytokine storms, is a life-threatening complication of COVID-19. During infection, the immune system is evoked to fight the pathogen. However, over-triggering the immune system also results in immunopathology. Hyperinflammation from COVID-19 might be triggered by the innate immune cells: macrophages, dendritic cells, and neutrophils, which are the first responders to infection, viral-induced pyroptosis, and decrease in type 1 interferon function or antibody-Fc receptor (FcR) interaction (reviewed in [76, 77]).

Anti-SAR-CoV-2 spike antibodies are involved in the activation of other immune cells or immune components through the Fc functions of antibodies, resulting in antibody-dependent cellular cytotoxicity, antibody-dependent cellular phagocytosis, complement-dependent cytotoxicity, and antibody-dependent cellular trogocytosis [78], which, if dysregulated, may progress to hyperinflammation [79]. Furthermore, the Fc of antibodies at suboptimal neutralizing concentration can introduce an extrinsic (classical) antibody-dependent enhancement (ADE) [80], another cause of uncontrolled inflammation. The neonatal Fc receptor (FcRn) retains immunoglobulin in the bloodstream, resulting in prolonged antibody responses. Therefore, antibodies targeting FcR and FcRn could be a therapeutic strategy to reduce the immunopathology of COVID-19 [81].

Interleukin-6 (IL-6) is a critical cytokine involved in this hyperactive immune response. It is a marker for COVID-19 progression and severity prognosis [82,83,84]; therefore, it is a target for controlling hyperinflammation. Several FDA-approved antibodies that block IL-6 and IL-6 receptor (IL-6R) interaction have been repurposed to treat COVID-19. Tocilizumab, a humanized anti-IL-6 receptor IgG1, was initially used to treat rheumatoid arthritis [85] and is the first monoclonal antibody approved for treating COVID-19 in hospitalized adults with severe COVID-19 [86]. Its effectiveness in improving the treatment outcome of COVID-19 treatment is controversial [84, 87]. Although no significant outcomes of tocilizumab were reported [88, 89], it was found to reduce disease severity and hospitalization time [84, 90,91,92]. However, no effect of tocilizumab on reducing COVID-19 mortality is inconclusive [90, 91, 93, 94]. Like tocilizumab, the role of sarilumab, the FDA-approved human anti-IL6R IgG1, in treating COVID-19 is controversial [95, 96] and requires further investigation.

Other proinflammatory molecules have been proposed as therapeutic targets [97,98,99,100,101]. Secukinumab, a monoclonal antibody against IL-17, the upstream IL-1 and IL-6 pathways, has been tested in a phase 2 clinical trial. There has been no improvement in the outcome of COVID-19 treatment, but a reduction of thromboembolism by secukinumab has been reported [98].

Combining the antibody with the inhibitor of the cytokine signaling molecule is another strategy to control hyperinflammation. A combination of secukinumab with baricitinib, a Janus kinase (JAK) 1 and 2 inhibitors, has shown benefits of reduction of ICU support and intubation, hospital stay, and lower mortality than treatment with baricitinib alone [102].

Besides directly targeting cytokines, nasal administration of anti-CD3 suppressed T cell function, reduced lung inflammation and serum IL-6, and increased TGFB1 expression [103].

Immunomodulation by inhibiting proinflammatory cytokines raises concerns about increased susceptibility to secondary infection [97, 104]. Treatment results are controversial [105], with either no effect reported on increasing the secondary infection [98, 102, 106, 108] or increasing the risk of secondary infection [107,108,109].

The effectiveness of immunomodulatory treatment for the recovery of dysregulated immune function in COVID-19 is multifactorial. The first is the administration time [84, 110], which may require calibration before dysregulation occurs to prevent the development of irreversible organ dysfunction [84, 111]. Individual factors and patient conditions also affect treatment outcomes (reviewed in [111]). Further research may focus on finding the best use of the treatment [110].

CD147

CD147 (EMMPRIN or basigin), a transmembrane glycoprotein in the immunoglobulin superfamily [109], has multiple binding partners to drive normal physiological functions and is involved in cancers and infectious diseases [112,113,114,115,116]. CD147 was reported as a receptor for SARS-CoV-2, which binds to the viral spike protein and facilitates viral entry of the cell lacking ACE2 [117•]. A humanized anti-CD147 antibody, meplazumab, was approved for phase I clinical trials for prophylaxis treatment for malaria and has been repurposed for the treatment of COVID-19 [117•]. Meplazumab reached the preclinical trial phase 2/3, effectively improving disease severity and mortality while reducing viral load and multiple cytokine levels [112, 118]. CD147 is involved in the viral entry process and plays a role in the inflammatory process [116, 119] and pulmonary fibrosis [120]. Therefore, blocking CD147 would help control infection and may mitigate the effect of cytokine-induced immunopathology and COVID-19 tissue fibrosis.

Engineered Antibody Format

Antibodies against COVID-19 were developed in different formats (Fig. 1), primarily as full-length antibodies with or without fragment crystallizable (Fc) engineering. Antigen-binding fragments (nanobody, single-chain antibody (scFv), Fab) can avoid Fc-induced ADE. They can be linked with other peptide/protein domains to improve the efficacy of antibodies or add an advantageous characteristic to the antibodies. Details of the format and designs of the engineered antibodies are described below.

Fig. 1
figure 1

Designs of SARS-Co-V2 therapeutic antibody format. A Conventional antibody with engineered Fc functions. B Antigen-binding fragments and antibody-fusion protein. C Multivalent/multispecific antibody

Full size image

Engineered Fc Antibodies

Knowledge of the interaction of the modified Fc has been long investigated with the in vitro and in vivo data and has already been approved for therapeutic products [121,122,123,124,125,126,127]. The engineered Fc for therapeutic antibodies for COVID-19 aims to (1) increase the half-life of the antibody and (2) decrease the immune activation and tissue injury caused by antibodies.

The neonatal Fc receptor (FcRn) is the first known receptor for transferring IgG from the mother to the fetus or the newborn [128]. Furthermore, FcRn plays a role in maintaining circulating immunoglobulin levels by binding and releasing IgG back into circulation. FcRn has been detected in epithelial, endothelial, and hematopoietic cells [129, 130]. The binding and release of Fc by FcRn are controlled by pH. Cells uptake IgG by pinocytosis, and IgG is entrapped into the endosome by FcRn. At low pH of the endosome, Fc binds to the FcRn and is sorted into tubules originating from the sorting endosomes to return to the plasma membrane [129, 131]. The increased pH (pH 7.4) causes a release of Fc [129, 131]. Binding to FcRn helps prevent IgG degradation and increases the serum half-life of the antibody. Several mutations increase the affinity of immunoglobulin molecules to FcRn or control the pH-dependent binding, resulting in prolonged circulating IgG levels. Additional details on the mechanisms and designs of the interaction between Fc and FcRn were reviewed in [129].

The functions of Fc are essential for viral clearance, reducing weight loss, and preventing the lethality of SARS-CoV-2 in animal models [79, 132, 133], and the defect in Fc function was related to the mortality of the COVID-19 patient [134]. However, Fc is not the only factor indicating the success of therapeutic antibodies for SARS-CoV-2 [136]. Antibodies with the Fc mutation, which affects FcR binding, demonstrated a therapeutic efficacy against COVID-19 [127, 137]. Stimulating the immune response by Fc function through FcR can induce a profound inflammatory response and ADE, leading to cytokine storms. Leucine positions 234 and 235, located in the CH2 domain of an antibody, and proline at position 329 or 331 are critical residues for Fc receptors and C1q binding. Mutations at these positions, such as LALA-PG (L234A/L235A/P329G) or TM (L234F/L235E/P331S), decreased the binding affinity of IgG1 to the Fc receptor and C1q molecule compared to the original [124, 125] and diminished the Fc effector function in vitro [125, 136]. The LALA (L234A/L235A) mutation also lowers the risk of Fc-mediated lung injury [27, 127]. Another way to reduce risk is to engineer Fc in an IgG4 isotype that cannot engage FcR [79, 138, 139].

Human anti-SARS-CoV-2 spike (RBD) antibodies, tixagevimab and cilgavimab (AZD7442), and etesevimab are examples of Fc-engineered antibodies for COVID-19. Tixagevimab and cilgavimab harbored the YTE (M252Y/S254T/T256E) and TM mutations to increase serum half-life (long-acting antibody) and reduce FcR and C1q complement binding, respectively [11, 137, 140]. Etesevimab contained the LALA mutation [141].

Besides engineered Fc, the half-life of the circulating antibody can also be prolonged by engineered variable regions of the antibody [142]. An engineered variable region with a lower molecular isoelectric point (pI) reduced antibody clearance in nonhuman primates [142]. The pI-engineered variable regions in combining the Fc mutation, N434A, which increased affinity for FcRn, were found in tocilizumab [129, 143], which is repurposed for treating COVID-19. Engineered Fc to increase activity to FcγRIIIa induced protective CD8 + T cell response against respiratory virus [144].

Nanobody (Single-Domain Antibody (sdAb))

The camelids have a particular type of antibody, i.e., heavy chain antibodies, which harbored only the heavy chain domain without the light chain counterpart [145, 146]. A unique characteristic of the antibody is the long CDR3, which helps bind to antigens to compensate for the lack of the light chain. The nanobody or single-domain antibody (sdAb) is a variable domain of the heavy chain antibody that functions in antigen binding. The size is ten times lower than conventional antibodies, making the molecules easy to express in a prokaryotic system and easy to manipulate and modify [36, 37, 39]. The nanobody is stable in harsh environments such as acidic, ionic strength, heat, proteolysis, and pH [39, 147,148,149,150].

Nanobodies against SARS-CoV-2 were developed, mainly against the S protein [34, 36, 37, 151, 152]. Long CDR3 and the small size of the nanobody may facilitate the single-domain antibody to the epitope that is hiding or is rarely targeted by conventional human antibodies [34, 36, 39, 152]. Another benefit of the nanobody is the lack of the Fc portion, reducing the risk of Fc-associated ADE [41]. However, enhanced virus infectivity by nanobodies was reported [151]. Nanobodies bound to enhancing epitopes on the RBD might induce conformational changes that promote interaction with receptors [151]. Nanobodies were also developed against nonstructural proteins [37, 67, 75]. The long CDR3 of the nanobody supports the accession to the cavity or enzymatic groove of the target [67, 69].

Bi-, Tri- and, Multivalent (Multispecific) Antibodies

Combining two or more antigen-binding domains, i.e., Fab, scFv, and nanobody of the antibody molecule, to increase the antibody’s avidity improved antibodies’ efficacies. Antigen-binding domains were linked together or with different molecules, commonly the Fc of IgG, to create the bi-, tri-, and multivalent antibody formats. These antibody formats also facilitated combinations of antigen-binding domains with different specificity to become bi-, tri-, or multispecific antibodies.

Fc-Supported Bi-, Trivalent Antibody

The scFv and nanobody have a small molecular weight, which helps tissue penetration and facilitates gene manipulation and fusion protein linkage but is rapid kidney clearance [78]. Linking the scFv or nanobody to Fc helped increase the half-life [47, 153] and assisted in the purification of the fusion proteins [154]. However, in some antibodies, linking scFv-Fc fusion to IgG affects neutralizing but not binding activity [155]. Fc-supported dimerization of molecules and demonstrated increasing avidity and improved efficiency compared to monovalent [34, 39, 41, 153].

The fusion of the antigen-binding domain with Fc also supports constructing bispecific and multivalent antibodies. Antigen-binding domains targeting different antigens can be combined by Fc dimerization to create the bispecific antibody. The bispecific antibody to different epitopes of the spike protein increased neutralization potency [156] and resistance to mutational escape [36, 59, 60]. Combining a neutralizing nanobody and a nonneutralizing Fab to spike protein improved antibody efficiency [157]. One or more antigen-binding domains can be added to the Fc at the N- and/or C-terminal to create a multivalent bispecific antibody format [158, 159]. For example, an anti-RBD spike linked to an IL-6 trans-signaling inhibitor (antibody to IL-6: IL6R complexes) prevents viral entry and cytokine release syndrome [76, 82, 154]. Unlike antibody cocktails, bispecific antibodies can reduce production costs and administration doses [156, 159]. Fc fusion and bispecific antibodies can be engineered to produce silent Fc to reduce the risk of ADE [35, 157].

Fc also facilitated the construction of multispecific trivalent antibodies [160], which improved the antibody’s potency and the prevention of viral escape. Interestingly, the molecules’ arrangement affected the antibody’s effectiveness [160]. Apart from fusion with the Fc of IgG, the decameric antibody was constructed by linking the antigen-binding domains to the Fc of the IgM, increasing the stability of the antibody for aerosolized administration to deliver the antibody directly to the lung [161].

Linker and Proteins Supported Multivalent (Multispecific) Antibody

The bivalent and trivalent antibody format can be generated by linking the molecules with the peptide linker [161,162,163]. The length of the linker can be a structurally guided design for the best potency [163]. The trivalent antibody format improved the stability of the antibody and is another format designed to be applied intranasally to function directly in the airway [161, 163].

Linking the antigen-binding domains of the antibody with the self-assembly protein or protein scaffold is another strategy for forming a multivalent antibody. Tetrameric antibodies are created by linking the Fab of scFv with or without Fc to the self-assembled human p53 tetramerizing domain, the best performance of the tetrameric molecule is the Ig format, which CH3 of the full-length antibody is linked to the p53 tetramerizing domain, and the hinge region of the antibody molecules was preserved [164]. With a similar principle, a multibody antibody (multispecific, multiaffinity) was developed using apoferritin, which can create a multimerization of 24 proteins [165]. Separately linking the different specificities of Fab and the Fc to apoferritin creates the multispecificity and multivalency complex that can be purified using protein A [165]. The multibody overcomes the point mutation and improves neutralization, even in nonneutralizing monovalent antibodies [165]. The attachment of nanobodies to lumazine synthase protein scaffold from Aquifex aeolicus, using a spy tag/spy catcher, creates a multivalent molecule with thermostability [166].

Antibody Fragment-Fusion Proteins and Other Antibody Formats

The antibody molecule usually targets extracellular antigens. In order to enable the function of the antibody within cells, scFv against the Nsp5 was linked to the cell-penetrating peptide (Cpp) as a superantibody [70]. The superantibody passed through the plasma membrane to inhibit viral replication.

The bivalent antibody linked to an antiviral peptide that blocked ACE2 binding was developed. The linker between the antibody and the peptide can be cleaved to separate the molecules at the site of action [162]. The antibody is PEGylation, commonly used to improve biological half-life and stability [162, 167].

A combination of 131I labeled antibodies for auger radiotherapy, electron energy penetrates deeply into the virus but not the nearby cells and can be applied for noninvasive imaging [168].

Conclusion and Perspective

Antibody therapy is considered an alternative treatment for COVID-19. Targeting and binding to multiple sites of the viral protein make the antibody-escaped mutation harder than the small-molecule drugs. Many therapeutic antibody design strategies have been developed to encounter SAR-COV-2 infection and complications. The challenges in antibody design are overcoming viral mutations and finding a therapeutic window for the antibody, particularly the immunomodulator [134]. Engineered antibodies with improved avidity and/or specificity were shown to be one strategy to avoid mutations. Multimeric antibody forms are stable to apply intranasally to function directly in the airway, reducing the concentration of antibodies and improving their effectiveness [169, 170]. Selection of the pancoronavirus conserved epitope(s) and using AI or computer-assisted designs of antigen–antibody interactions, intermolecular linkage, and immune escape mechanism predictions would help develop therapeutic antibodies [163, 171,172,173]. The benefits and difficulties of the engineered antibodies are summarized in Table 1. Apart from designing the paratope, fast isolation and efficient production of therapeutic antibodies are other factors that need cohesive development to make a successful therapeutic antibody.

Table 1 Benefits and difficulties of antibodies as well as engineered antibody formats
Full size table