Bispecific antibodies are the hybrids of two different antibody molecules and can simultaneously bind to two different types of antigens (Kontermann, R. E., & Brinkmann, 2015). They can be constructed by chemically cross-linking two different antibodies or synthesized by the hybridomas formed by fusing two different monoclonal-antibody-producing cell lines. These methods generate mixtures of monospecific and bispecific antibodies from which the desired bispecific molecule is purified. Bispecific antibodies can also be generated using genetic engineering techniques to construct genes that encode molecules only with the two desired specificities. Therefore, when bispecific antibodies are injected into the body, they can simultaneously bind to two different types of cells (Fig 1).

For instance, in a bispecific antibody used in cancer treatment, one-half of the antibody has specificity for a cancer cell. The other half of Ab has specificity for the CD3 present on cytotoxic T cells (Tc cells). Additionally, the Fc region of bispecific antibodies can bind to the immune effector cells that express Fc receptors, such as an NK cell, an activated macrophage, a dendritic cell, etc. (Fig 2). The net effect is that the bispecific antibody links Tc cells and Fc receptor-expressing effector cells like macrophages, NK cells, dendritic cells to the cancer cells (Krah et al., 2018). Activated Tc cells kill the cancer cells by producing proteins like perforin and granzyme that enter the cancer cell and initiate the cell’s apoptosis. On the other hand, the effector cells mediate the destruction of the cancer cell through the Antibody-dependent cellular toxicity or ADCC process.

Bispecific antibodies are more potent (more than 1,000-fold) in eliminating and killing tumor cells than conventional antibodies because of their ability to recruit both cytotoxic T cells and effector cells simultaneously to the tumor cells to enhance their killing. On the other hand, the conventional antibodies don’t activate T cells because their Fab regions are already used for binding the tumor cells. Thus, the bispecific antibodies might outshine the conventional mAbs as cancer therapeutics.

Types of bispecific antibodies

There are two common formats of bispecific antibodies available: the single-chain variable fragment (scFv)-based antibody (no Fc fragment) and the full length IgG-based antibody (Wang et al., 2019).

Single-chain variable fragment (scFv)-based antibodies

The scFv based antibody is produced by fusing variable domains of the IgG heavy chain (VH) and the light chain (VL) through a short flexible peptide linker of about 10–25 amino acids (Ahmad et al., 2012; Monnier et al., 2013). For the generation of the scFv construct, mRNA is isolated from the hybridoma cell that produces mAbs against the desired antigen. mRNA is then reverse transcribed into cDNA. After this step, using specific primers, Polymerase chain reaction, abbreviated as PCR, amplifies the DNA sequence corresponding to the antibody VH and VL genes.

The scFv gene construct is designed in the orientation, VL-linker-VH (Fig 3).

It is then either expressed in microbial systems like E. coli or in mammalian systems such as HEK293 cells. In the expressed scFv, the C-terminus of the VH is linked to the N-terminus of the VL through the peptide linker.

In short, scFv is the smallest unit of the immunoglobulin molecule with a molecular weight in the range of 25kDa and functions in antigen-binding activities. Since an scFv molecule contains one VH and VL domain, it has a single antigen-binding site. Additionally, each VH and VL domain of scFv contains three complementarity-determining regions (CDRs). The combination of the CDRs of the VH and the CDRs of the VL determines the binding specificity of the scFv molecule.

Also, the scFv-based antibodies are the smallest recombinant antibodies (rAb) that retain the antigen-binding activity and are popularly used for therapeutic and analytical applications because of their small size and the possibility of genetic engineering.

In addition, glycine, serine, and threonine are preferred amino acids for designing peptide linkers because their short side chains grant conformational flexibility and minimal immunogenicity to the scFv molecules. Also, serine and threonine amino acids improve the solubility of scFv molecules.

There are three types of scFv-based antibodies: bispecific T-cell engager (BiTE), dual-affinity re-targeting proteins (DARTs) and Tandem diabodies (TandAbs).

Bispecific T-cell engager (BiTE)

This type of bispecific antibody is a fusion protein containing two scFv fragments from two different monoclonal antibodies, which are connected by a peptide linker (Fig 4). The resulting ∼ 55-kDa molecule is a single polypeptide with N- and C-terminus. The two scFv fragments (Fv-A and Fv-B) retain each antibody’s binding activity when assembled (Slaney et al., 2018; Tian et al., 2021).

The linker connecting the two scFvs enables their free rotation, vital for flexible interaction with targeted receptors on two target cells. For instance, one scFv fragment binds to cytotoxic T-cell, and another scFv fragment binds to the cancer cell (Fig 5). Simultaneously binding to both T cell and cancer cell induces T-cell activation and subsequently targets the T-cell’s cytotoxicity activity against the cancer cell, ultimately killing it by producing proteins like perforin and granzyme that enter the cancer cell and initiate the cell’s apoptosis.

One example of BiTE drugs is blinatumomab, which has been approved by the FDA for the treatment of relapsed or refractory acute lymphoblastic leukemia (ALL) in adult patients. Blinatumomab consists of one anti-CD19 scFv connected through a short glycine–serine linker to the second anti-CD3 scFv. Thus, the drug links CD-3 containing T cells with CD-19 receptors found on the surface of B cell lymphoma (Wu et al., 2015; Goebeler et al., 2016).

Another BiTE drug is Solitomab which is under clinical trials. This drug also consists of two scFv fragments, in which one of the scFvs binds to T cells via the CD3 receptor and the other to EpCAM antigen, which is expressed by gastrointestinal, lung, prostate, ovarian, and other cancer cells (Bellone et al., 2016; Kebenko et al., 2018).

Dual-affinity re-targeting proteins (DARTs)

The linker sequences that connect the V regions in the BiTE structure constrain the necessary conformational flexibility required during antibody-antigen recognition, reducing the binding efficiency. To address this issue, DART proteins were developed.

A DART molecule consists of two engineered scFv fragments which have their VH and VL domains exchanged. Precisely, the scFv1 consists of a VH from antibody B and a VL from antibody A, while the scFv2 consists of VH from Ab-A and VL from Ab-B (Fig 6). Thus, when scFv1 and scFv2 heterodimerize, they form a bispecific DART molecule with two unique antigen-binding sites (Wang et al., 2019).

This configuration of DART molecules lacks the constraint of an intervening peptide linker, allowing DART to mimic the natural interaction within an IgG molecule. In addition, adding a cysteine residue to the end of each heavy chain improves the stability of the bispecific DART molecule by forming a C-terminal disulfide bridge.

A study by Moore et al. on comparing in vitro ability of CD19xCD3 DART and BiTE molecules showed that DART molecules were better at redirecting CD3 expressing T cells to kill CD19 expressing B-cell lymphoma. In this study, both DART and BiTE molecules were derived from the same parental antibodies (mouse anti-human CD3 and CD19 mAbs).

Tandem diabodies (TandAbs)

The small size of scFvs BiTE and DART contributes to a high renal clearance rate in comparison to natural IgG antibodies. The size issue can be resolved by generating Tandem diabodies (TandAbs). A tandem diabody is formed by non-covalent association of tandem two-hybrid scFv fragments consisting of VH and VL domains of different specificities (say antigen A and B) (Fig 7).

In addition, the tandem diabodies have tetravalent properties, which means they provide two binding sites for each antigen. Thus, they are also called tetravalent bispecific antibodies (Kipriyanov, 2009; Azhar et al., 2017).

Moreover, TandAbs have a molecular weight of approximately 105 kDa that exceeds the renal clearance threshold (60–70kDa), thus offering a longer half-life than smaller antibody constructs. Two TandAb bispecific drugs are in clinical trials — AFM13 (CD30xCD16) and AFM11 (CD19xCD3).

Role of the linker in designing scFv-based antibodies

Two essential features to be considered while designing the linker are amino acid composition and sequence length (Gu et al., 2010; Stamova et al., 2012; Yusakul et al., 2016).

(i) Amino acid composition is critical in designing a flexible linker peptide. The most commonly used amino acid sequence motif in designing linker is (G4S)n (G: glycine, S: serine; G4S means four glycine residues and one serine residue) (Fig 8). Glycine and serine are preferred amino acids because their short side chains grant conformational flexibility and minimal immunogenicity, while serine improves solubility. Besides the Gly-Ser linker, other charged amino acid residues such as glutamic acid and lysine can also be used to enhance solubility.

(ii) Length of the linker between variable domains of the heavy and light chain is also critical in designing scFv-based antibodies. The success of scFv-based antibodies construction largely depends on the peptide linker between the VH and VL domains. The linker should neither interfere with the folding and association of the VH and VL domains nor reduce stability and binding activity.

The length of the linker has a significant impact on the monomer or multimer formation of the antibodies. For instance, a linker length longer than 12 amino acid residues allows sufficient distance between heavy and light-chain domains to associate and form monomers. On the other hand, shorter linkers connecting VH and VL domains prevent the direct association of the two domains to create functional Fv domain, rather result in an increased possibility for pairings between VH and VL of different scFv molecules, forming dimers, trimers, or multimers. Therefore, for the successful construction of bispecific scFv, it is important to optimize linker length.

For instance, in the case of BiTE scFv-based antibody, long linkers are placed between the homologous light chain and heavy chain variable domains to ensure association to form a functional Fv. On the other hand, short linkers are placed between heterologous heavy-chain fragments to create the connection between the two Fvs (Fig 9a).

In the case of the DART molecules that bind to each other to form functional dimers, the linkers between VHA and VLB or VHB and VLA need to be as short as five amino acids to prevent their association from creating an undesired Fv (Fig 9b). Instead, because of a short linker, one scFv1 molecule associates with another scFv2 molecule (heterodimerization) to form a bivalent scFv-based antibody. Moreover, the positioning of the disulfide bond is another critical feature of DART molecules, which holds the molecule together in the correct orientation.

And in the case of TandAb, the linkers between adjacent domains are too short (three amino acids) for the two variable regions to fold together, forcing scFvs to dimerize and form diabodies. Conversely, the linker length of less than three amino acids (one or two amino acids) favors the formation of trimers or tetramers, so-called triabodies or tetrabodies.

Full-length IgG-based antibody

Although IgG-like asymmetric bispecific antibodies have certain properties similar to natural monoclonal IgG antibodies, they are engineered molecules that have not been generated by typical plasma cells. As a result, these differences lead to significant production challenges. One of the greatest challenges for generating asymmetric IgG-like bispecific antibodies is ensuring the correct assembly of antibody fragments (Grote et al., 2012; Fan et al., 2015). A random assembly of four distinctive polypeptide chains (two different heavy and two different light chains) results in 16 combinations, among which only two represent the desirable asymmetric heterodimeric bispecific antibody. The remaining pairings result in non-functional or monospecific molecules. Therefore, two key things are required to be considered while producing the desired IgG-like bispecific antibody — the heterodimerization of two different heavy chains and the discrimination between the two light-chain/heavy-chain interactions. Full-length IgG-based antibodies can be generated by different strategies. For instance,

Quadroma (or Hybrid-Hybridoma) Technology

Bispecific antibodies can be generated by the hybridomas that are formed by the somatic fusion of two different hybridoma cells (Fig 10). Each hybridoma cell expresses a unique monoclonal antibody with predefined specificity. So, when the two different antibody-expressing hybridoma cells are fused, the resulting hybrid-hybridoma cell produces the immunoglobulin heavy and light chains from both parent hybridoma cells, and their assembly allows the formation of bispecific antibodies.

Suppose 2 H chains from both parent hybridoma cells (aAAa and aBBb) are represented as A and B, and light chains are represented as a and b. The random assembly of two different heavy and two different light chains can theoretically result in (2⁴) 16 combinations, and only one of those is a functional bispecific antibody (Fig 11). Six of these combinations occur twice due to symmetry and produce similar antibodies (aABb, aBBb, aABa, bABb, aAAb, and bABa). The remaining four combinations (aAAa, bAAb, bBBb and aBBa), occur only once.

Thus, ten types of antibodies are produced. Six of them are bispecific antibodies, out of which only one is a functional bispecific antibody and the resulting combinations result in non-functional antibodies.

Catumaxomab (anti-EpCAM x anti-CD3) is the first approved IgG-like bispecific antibody generated via quadroma technology by the fusion mouse IgG2a producing hybridoma cell and rat IgG2b producing hybridoma cell.

In Catumaxomab antibody, one Fab antigen-binding site binds T-cells via the CD3 receptor, and the other site binds tumor cells via the tumor antigen epithelial cell adhesion molecule (EpCAM). On the other hand, the Fc region of the antibody recruits and activate immune effector cells via binding to FcγR receptors present on NK cells, dendritic cells, activated macrophages, etc. (Seimetz et al., 2010; Heiss et al., 2010).

However, immunogenicity remains the concern with Catumaxomab antibody, a chimeric mouse/rat bispecific Ab. Because when injected in patients, they may generate human anti-mouse or anti-rat antibody responses, which can lead to the negation of Catumaxomab’s drug-related effects, thus completely inhibiting the therapeutic aspect of the drug. The anti-drug antibodies may also cause allergic and several other adverse effects in the patients. Therefore, it is essential to humanize antibodies as much as possible to lower the generation of anti-drug-antibodies in the patients.

Knobs-into-holes (KIH) technology

The second method of generating bispecific antibodies is Knobs-into-holes technology (KIH), which refers to producing either a “knob” or a “hole” in the CH3 domain of the Fc region of each heavy chain to promote heterodimerization. The knobs and holes are engineered in the CH3 domain because H chains of an IgG antibody directly interact at their CH3 domains (Fig 12). However, at CH2 domains, they interact through the carbohydrate bound to the asparagine (Asn). Therefore, several amino acid changes must be done to make a knob on the CH3 domain of the heavy chain of mAb1 and a hole on the CH3 of the heavy chain of mAb2 (Xu et al., 2015).

To generate a knob, smaller amino acid in the CH3 domain of the H chain of mAb1 is replaced with a larger amino acid; for instance, threonine (T) at position 366 is replaced with larger amino acid tyrosine (Y). On the other hand, to generate a hole, the larger amino acid in the CH3 domain of the H chain of another mAb2 is replaced with a smaller amino; for instance, tyrosine at position 407 is replaced with a smaller amino acid threonine (Wang et al., 2019).

The engineered CH3 domains containing knob and hole fit into each other, favoring the heterodimerization. The heteromeric heavy chains have the ability to produce functional bispecific antibodies and also retain Fc-mediated effector functions, such as ADCC and complement activation. An identical light chain is used for each arm of the IgG-based bispecific antibody. Therefore, coexpression of engineered heavy chains with the light chains either in E. coli or mammalian cells, e.g., human embryonic kidney (HEK) cells, produce the IgG-based bispecific heterodimer with two antigen-binding sites (Rouet and Christ, 2014).

CrossMab technology

The Knobs-into-holes technology enables correct heavy-chain heterodimerization and generates desired bispecific antibodies when identical light chains are used. However, it is not always possible to use a common light chain because, in many cases, antigen recognition critically relies on the partner light chain. By using two different light chains, their random assembly with the heterodimer heavy chain (obtained by KIH technology) results in the generation of four different combinations of antibodies, with only one being the desired bispecific antibody (Fig 13).

To solve the problem of light chain mispairing, CrossMab technology has been developed that allows the correct pairing of the light chain and heavy chain to minimize the generation of non-functional and undesired bispecific antibodies (Liu et al., 2017).

The basis of the CrossMab technology is the crossover of domains of the light and heavy chains within one arm of a bispecific IgG antibody enabling correct light chain association. In contrast, correct heterodimerization of the heavy chains is achieved by the knob-into-hole technology (KIH). CrossMab is achieved in three ways (Schaefer et al., 2011; Klein et al., 2019):

1. In the first method, the entire heavy chain of the Fab region of one-half of the bispecific antibody is interchanged with a cognate light chain to generate CrossMab Fab bispecific Ab (Fig 14). The “crossover” obtained retains the binding affinity while favoring the assembly of the engineered Fab fragment.

2. The second method involves swapping the VH of a Fab region with its corresponding VL domain to generate CrossMab VH-VL bispecific Ab (Fig 15).

3. In the third method, CH1 and CL of one Fab region are interchanged to generate CrossMab CH1-CL bispecific Ab (Fig 16).

Such crossover designs reduce the heavy–light chain mispairing to yield functional bispecific antibodies with the potential to recognize two antigens simultaneously. Additionally, the cross-over bispecific antibodies retain the affinity and stability of the parent antibodies.

For instance, Vanucizumab (CrossMabCH1-CL) bispecific antibody exhibits antitumor activity by targeting vascular endothelial growth factor (VEGF-A) and angiopoietin-2 (Ang-2) simultaneously (Fig 17). As a result, VEGF-A and Ang-2 are unable to bind to their receptors VEGFR and Tie-2, respectively, because of which angiogenesis in tumor cells is hampered (Hidalgo et al., 2018; Heil et al., 2021).

Angiogenesis is a highly regulated process by which tumors develop new vasculature because for cancer cells to proliferate, a continuous supply of oxygen is needed. But without the formation of new tumor vasculature in the presence of Vanucizumab, tumor growth is arrested.

References

Ahmad, Z. A., Yeap, S. K., Ali, A. M., Ho, W. Y., Alitheen, N. B. M., & Hamid, M. (2012). scFv antibody: principles and clinical application. Clinical and developmental immunology, 2012.

Azhar, A., Ahmad, E., Zia, Q., Rauf, M. A., Owais, M., & Ashraf, G. M. (2017). Recent advances in the development of novel protein scaffolds based therapeutics. International journal of biological macromolecules, 102, 630–641.

Bellone, S., Black, J., English, D. P., Schwab, C. L., Lopez, S., Cocco, E., … & Santin, A. D. (2016). Solitomab, an EpCAM/CD3 bispecific antibody construct (BiTE), is highly active against primary uterine serous papillary carcinoma cell lines in vitro. American journal of obstetrics and gynecology, 214(1), 99-e1.

Fan, G., Wang, Z., Hao, M., & Li, J. (2015). Bispecific antibodies and their applications. Journal of hematology & oncology, 8(1), 1–14.

Goebeler, M. E., Knop, S., Viardot, A., Kufer, P., Topp, M. S., Einsele, H., … & Bargou, R. C. (2016). Bispecific T-cell engager (BiTE) antibody construct blinatumomab for the treatment of patients with relapsed/refractory non-Hodgkin lymphoma: final results from a phase I study. Journal of Clinical Oncology, 34(10), 1104–1111.

Grote, M., Haas, A. K., Klein, C., Schaefer, W., & Brinkmann, U. (2012). Bispecific antibody derivatives based on full-length IgG formats. In Antibody Methods and Protocols (pp. 247–263). Humana Press, Totowa, NJ.

Gu, X., Jia, X., Feng, J., Shen, B., Huang, Y., Geng, S., … & Long, M. (2010). Molecular modeling and affinity determination of scFv antibody: proper linker peptide enhances its activity. Annals of biomedical engineering, 38(2), 537–549.

Heil, F., Babitzki, G., Julien-Laferriere, A., Ooi, C. H., Hidalgo, M., Massard, C., … & Lechner, K. (2021). Vanucizumab mode of action: Serial biomarkers in plasma, tumor, and skin-wound-healing biopsies. Translational Oncology, 14(2), 100984.

Heiss, M. M., Murawa, P., Koralewski, P., Kutarska, E., Kolesnik, O. O., Ivanchenko, V. V., … & Parsons, S. L. (2010). The trifunctional antibody catumaxomab for the treatment of malignant ascites due to epithelial cancer: Results of a prospective randomized phase II/III trial. International journal of cancer, 127(9), 2209–2221.

Hidalgo, M., Martinez-Garcia, M., Le Tourneau, C., Massard, C., Garralda, E., Boni, V., … & Krieter, O. (2018). First-in-human phase I study of single-agent vanucizumab, a first-in-class bispecific anti-angiopoietin-2/anti-VEGF-A antibody, in adult patients with advanced solid tumors. Clinical Cancer Research, 24(7), 1536–1545.

Kebenko, M., Goebeler, M. E., Wolf, M., Hasenburg, A., Seggewiss-Bernhardt, R., Ritter, B., … & Fiedler, W. (2018). A multicenter phase 1 study of solitomab (MT110, AMG 110), a bispecific EpCAM/CD3 T-cell engager (BiTE®) antibody construct, in patients with refractory solid tumors. Oncoimmunology, 7(8), e1450710.

Kipriyanov, S. M. (2009). Generation of bispecific and tandem diabodies. In Antibody Phage Display (pp. 177–193). Humana Press.

Klein, C., Schaefer, W., Regula, J. T., Dumontet, C., Brinkmann, U., Bacac, M., & Umaña, P. (2019). Engineering therapeutic bispecific antibodies using CrossMab technology. Methods, 154, 21–31.

Kontermann, R. E., & Brinkmann, U. (2015). Bispecific antibodies. Drug discovery today, 20(7), 838–847.

Krah, S., Kolmar, H., Becker, S., & Zielonka, S. (2018). Engineering IgG-like bispecific antibodies — an overview. Antibodies, 7(3), 28.

Liu, H., Saxena, A., Sidhu, S. S., & Wu, D. (2017). Fc engineering for developing therapeutic bispecific antibodies and novel scaffolds. Frontiers in immunology, 8, 38.

Monnier, P. P., Vigouroux, R. J., & Tassew, N. G. (2013). In vivo applications of single chain Fv (variable domain)(scFv) fragments. Antibodies, 2(2), 193–208.

Moore, P. A., Zhang, W., Rainey, G. J., Burke, S., Li, H., Huang, L., … & Johnson, S. (2011). Application of dual affinity retargeting molecules to achieve optimal redirected T-cell killing of B-cell lymphoma. Blood, The Journal of the American Society of Hematology, 117(17), 4542–4551.

Rouet, R., & Christ, D. (2014). Bispecific antibodies with native chain structure. Nature biotechnology, 32(2), 136–137.

Schaefer, W., Regula, J. T., Bähner, M., Schanzer, J., Croasdale, R., Dürr, H., … & Klein, C. (2011). Immunoglobulin domain crossover as a generic approach for the production of bispecific IgG antibodies. Proceedings of the National academy of Sciences, 108(27), 11187–11192.

Seimetz, D., Lindhofer, H., & Bokemeyer, C. (2010). Development and approval of the trifunctional antibody catumaxomab (anti-EpCAM× anti-CD3) as a targeted cancer immunotherapy. Cancer treatment reviews, 36(6), 458–467.

Slaney, C. Y., Wang, P., Darcy, P. K., & Kershaw, M. H. (2018). CARs versus BiTEs: a comparison between T cell–redirection strategies for cancer treatment. Cancer discovery, 8(8), 924–934.

Stamova, S., Koristka, S., Keil, J., Arndt, C., Feldmann, A., Michalk, I., … & Bachmann, M. (2012). Cancer immunotherapy by retargeting of immune effector cells via recombinant bispecific antibody constructs. Antibodies, 1(2), 172–198.

Tian, Z., Liu, M., Zhang, Y., & Wang, X. (2021). Bispecific T cell engagers: an emerging therapy for management of hematologic malignancies. Journal of Hematology & Oncology, 14(1), 1–18.

Wang, Q., Chen, Y., Park, J., Liu, X., Hu, Y., Wang, T., … & Betenbaugh, M. J. (2019). Design and production of bispecific antibodies. Antibodies, 8(3), 43.

Wu, J., Fu, J., Zhang, M., & Liu, D. (2015). Blinatumomab: a bispecific T cell engager (BiTE) antibody against CD19/CD3 for refractory acute lymphoid leukemia. Journal of hematology & oncology, 8(1), 1–7.

Xu, Y., Lee, J., Tran, C., Heibeck, T. H., Wang, W. D., Yang, J., … & Yin, G. (2015, January). Production of bispecific antibodies in “knobs-into-holes” using a cell-free expression system. In MAbs (Vol. 7, №1, pp. 231–242). Taylor & Francis.

Yusakul, G., Sakamoto, S., Pongkitwitoon, B., Tanaka, H., & Morimoto, S. (2016). Effect of linker length between variable domains of single chain variable fragment antibody against daidzin on its reactivity. Bioscience, biotechnology, and biochemistry, 80(7), 1306–1312.