In cancer, numerous tumor associated antigens (TAAs) have been utilized as disease markers and targets for antibody therapeutics, for example p53, HER-2, EGFR, among others. Monoclonal antibodies have been the most powerful tools to dissect the expression and to block the functions of tumor associated antigens. Several monoclonal antibody products have been developed for cancer diagnosis and treatment. The market for therapeutic monoclonal antibodies is forecasted to reach ∼$ 94 billion by 2017 and nearly $ 125 billion by 2020. The antibody market has yet reached its full potential. This is because relative few targets are being exploited. Up to this date, there has been little systematic effort to discover novel tumor associated antigens and develop therapeutic antibody leads through large scale generation of anti-tumor monoclonal antibodies.

Recently, BostonMolecules has developed rabbit B cell derived monoclonal antibody (RbcMAbs) technology which provides a novel type of monoclonal antibody with higher affinity, higher sensitivity and specificity. In addition, a single immunized rabbit can generate a large panel of high affinity antibodies that recognize many different epitopes, compared to traditional mouse monoclonal technology. BostonMolecules has also developed a technology platform to subtractive-immunize rabbits with disease tissues (including tumor tissue) and subsequently generate large panel of disease specific RbcMAbs. In addition, the company recently succeeded to make the first humanized RbcMAbs, which paved the way to therapeutic development.

The disease specific RbcMAbs platform allows to generate a large anti-tumor RbcMAbs bank and subsequently screen for functional therapeutic antibodies. It consists of 5 stages: (1) Subtractive-immunization of rabbits with tumor tissues; (2) Generation of a large RbcMAbs bank; (3) High throughput screening of RbcMAbs for therapeutic leads using cell-based assays; (4) Humanization of RbcMAbs leads, and (5) animal models and preclinical characterization. It will be the first systematic effort to screen therapeutic antibodies using the superior rabbit antibody repertoire. Through this platform, scientistscan discover novel cancer targets, especially membrane-bound TAAs, discover useful RbcMAbs for future diagnostic development and develop valuable therapeutic antibodies ready for clinical trials.

Tumor-associated antigens
The rationale for the use of antibody-based agents for the diagnosis and treatment of cancer is based the fact that cancerous cells present antigenic moieties that are not detectable on cells in normal tissues. Only a handful of antigens are truly tumor-specific, such as clone-specific idiotypic (Id) immunoglobulins. Most proteins are not strictly tumor-specific; rather, they have restricted distribution on normal tissues. Typically, so-called “tumor antigens” are present and accessible only in small amounts outside the target tumor tissue. These antigens are called tumor-associated antigens (TAAs).

Traditionally, TAAs have been grouped into two categories: 1) TAAs defined by T cells, and 2) TAAs defined by antibodies. MAGE-1 was discovered as the first human TAA by a gene-cloning method for the identification of antigens by autologous human cytolytic T cells (van der Bruggen et al., 1991). Most T cell-defined human TAAs were first found in melanomas, including the MAGE, BAGE, GAGE and RAGE group of antigens. The method of identifying this type of TAAs include: 1) biochemical characterization of peptides eluted from the MHC I molecules on tumor cells (Cox et al., 1994), and 2) the in vitro evaluation of T cell-stimulating activity of peptide antigens that are encoded by genes overexpressed or mutated in the tumor under investigation (Jung S, 1991).

Antibody-defined TAAs have traditionally been divided into oncofetal antigens, e.g. carcinoembryonic antigen (CEA), breast cancer mucin antigen (MUC1), alpha fetoprotein (AFP) and prostate-specific membrane antigen (PSMA) (Baldus and Hanisch, 2000; Elgamal et al., 2000; Hammarstrom, 1999); oncogene products, e.g., nonmutated HER-2/neu (Ross and Fletcher, 1999) and p53 (Kaelin, 1999) gene products; tissue-lineage antigens (also called tissue-specific antigens) which continue to be expressed on cancer cells (Ghose and Blair, 1987), e.g. prostate-specific antigen (PSA), tyrosinase, and gp100.

The recent use of the SEREX approach (serological identification of antigens by recombinant expression cloning) has revealed that human tumors express multiple antigens that can elicit an antibody response in the autologous host (Pfreundschuh et al., 1978; Sahin et al., 1997). Therefore, a high throughput immunosubtraction method to systematically screen for tumor reacting rMAbs, will become a valuable approach for discovering novel TAAs.

Subtractive immunization

MAbs are a powerful tool for discovering differentially expressed proteins such as TAAs. A common approach for developing MAbs is to immunize mouse with tumor cells, cell lysates and other complex antigens. A major problem with this method is that MAbs are produced predominantly to the dominant antigens. However, many TAAs are likely less abundant and thus can not be discovered by direct immunization with tumor cells.

Subtractive immunization was developed in the mouse system to increase the chances of producing antibodies to low abundant but functionally relevant antigens. One way of achieving this goal is through the use of the immunosuppressive agent cyclophosphamide coupled to immunizations with two distinct cell variants as sequential immunogens (Brooks et al., 1993; Hanski et al., 1991; King and Morrow, 1988; Sensenbrenner et al., 1979; Williams et al., 1992).

Cyclophosphamide selectively kills activated B cells that have been stimulated to proliferate (Many and Schwartz, 1970). It was used to partially suppress the immune system in mice to decrease the reactivity to common cell surface antigens present on a non-metastatic cell variant that was used as the initial immunogen (Brooks et al., 1993). Cyclophosphamide has also been used in suppressive immunization procedures to make MAbs against antigens on closely related neuronal cells (Matthew and Sandrock, 1987) and to specific antigens on tumor cells (Shestowsky et al., 1990). Using cycles of cyclophosphamide treatment, mouse MAbs were generated against cryptic epitopes in denatured and proteolyzed collagens (Xu et al., 2000). This compound has been shown to induce stronger immune tolerance than a neonatal tolerization method (Williams et al., 1992)

Other methods have been invoked in attempts to suppress the immune response for dominant epitopes. Passive immunization and successive antigen depletion, using E. coli cytoplasmic proteins as antigenic mixture were tested in rabbits, and shown to be partially effective (Thalhamer and Freund, 1984; Thalhamer and Freund, 1985). Epitope-masking was attempted for human cell-surface antigens to make mouse MAbs (Su et al., 1996). Rabbit antiserum from tester cells that are preabsorbed by driver cells was used as a probe to clone adipocyte-specific proteins (Scherer et al., 1998). None of these approaches have been tested using mammalian cells and lysates for subtractive immunization of rabbits.

Rabbit monoclonal antibody

Most of antibodies approved for clinical area are mouse origin. However, the mouse system is limited by a small spleen and the mice used are usually inbred, thereforeproviding a less diversity of immune responses. In contrast, the rabbit has a robust immune system and bigger spleen to generate antibodies with high affinity and specificity.Traditionally, it has been difficult to raise monoclonal antibodies in the rabbit through immortalized B lymphocytes by fusing with myeloma cells. Recombinant antibody approaches were reported that rabbit monoclonal antibody genes were isolated from single cell PCR or phage display (Babcook, Leslie et al. 1996; Rader, Ritter et al. 2000). However, its general application is limited by the difficulty of single cell selection and PCR, or by technical demanding protocols of phage display. Current advance in Rabbit B-cell Cloning MAbs technology (RbcMAbs) provide a robust platform to generate rabbit MAbs through combination of Single B-cells proliferating technology with recombinant antibody cloning technology.

The availability of rabbit monoclonal antibodies is highly desirable. Comparing withMAbs with mouse origin in laboratory use, RbcMAbs are demonstrating their potential for clinical applications by offering many advantages over mouse MAbs. Compared with antibodies from other sources, RbcMAbs has at least following advantages: higher binding affinity,wider repertoire, simpler structure, robust reproduction and easy to be humanized.

First, antisera from rabbits are generally of higher affinity than the equivalents in mice. Second, rabbit monoclonal antibodies are expected to recognize many antigens that are not immunogenic in mice, including mouse proteins (Krause 1970; Bystryn, Jacobsen et al. 1982; Norrby, Mufson et al. 1987; Weller, Meek et al. 1987; Raybould and Takahashi 1988). For example, Bystryn et al (Bystryn, Jacobsen et al. 1982) compared rabbit and mouse antibodies directed against human melanoma cells and showed that they recognize different epitopes. Our own data also indicated that rabbit recognize more epitopes of digested human fibronectin than mouse (data not shown). Third, Because of the size of the rabbit spleen, more antigen specific B cells can be selected, making MAbs generation at industrial scale a feasible task. Fourth, BIAcore analysis of randomly selected 12 rabbit MAbs showed that the affinity is within the range of 10-9 to 10-12 M, much higher than average mouse MAbs. Fifth, thousands of single B cells can be expanded and RbcMAbs generated from each immunized spleen, providing a much greater number of independent monoclonal antibodies that recognize different epitopes. The antibody containing B cell supernatant can be directly used for drug screen without need to remove HAT from the medium as traditional mouse hybridoma technology should. Thus, a panel of bioactive RbcMAbs could be easily obtained for further selection of antibody drug leads. The robustness in the generation of RbcMAbs offers a much higher success rate to get the most desirable drug leads in a relatively shorter period of time. Sixth, RbcMAbs are easy to be humanized.

A novel humanization technology termed Mutational Lineage-Guided (MLG) humanization was recently developed to humanize RabMAb much easier. MLG humanization is based on the fact that amino acid sequences of the variable regions of the heavy and light chains (VH and VL) from the collection of IgG sequence are aligned to form a phylogenetic tree. Related antibodies are grouped according to their sequences similarity to each other. Conserved sequences in a lineage-related group represent residues critical to the structure and function of IgG while uncon-served residues have less even no effects on the biological activities. Since these variable positions were obtained from a group of antibodies usually from one parental B cell, they must have been effectively examined by an animal immune system. Thus, substitution of amino acids at these positions to humanize antibodies should be well tolerated without sacrificing antibody specificity and affinity. More importantly, such variations are found not only in the framework regions, but also in the CDRs. Therefore, MLG humanization can be applied to the humanization of the framework regions as well as the CDRs. Due to the large numbers of lymphocytes of rabbit spleen, enough bioactive RbcMAbs will be available to generate humanized RbcMAb by MLG humanization.

Functional antibodies

Membrane-bound proteins are an important class of proteins that include receptors, transporters, adhesion molecules, and channel proteins. Functional study of membrane-bound human proteins has significant value in understanding many human diseases. Membrane proteins represent one of the most promising classes of therapeutic targets. The genomic initiative has discovered a large number of potential membrane proteins through algorithms that recognize a stretch of hydrophobic amino acids. Such an approach to identify potential membrane proteins by sequence analysis has a limitation in accuracy and typically requires knowledge of the entire coding sequence, including the N terminus (Nielsen, Brunak et al. 1999), which is not currently available for most of the human genes.

Several experimental approaches for identifying secreted and membrane proteins have been described, but each method has its own limitation and none of the methods focused on membrane proteins only (Tashiro, Tada et al. 1993; Klein, Gu et al. 1996; Zannettino, Rayner et al. 1996; Kopczynski, Noordermeer et al. 1998; Scherer, Bickel et al. 1998; Diehn, Eisen et al. 2000). MAbs generated against membrane proteins are powerful tools for basic research and can become potential diagnostic and therapeutic leads. Several mechanisms account for the therapeutic effects of therapeutic antibodies directed to specific membrane targets: functional blocking/stimulating , antibody dependent cytotoxicity (ADCC), complement dependent cytotoxicity (CDC), activation by cross-linking receptors. In addition, conjugated antibodies exert therapeutic effects through the payload that antibodies carry, such as toxins, cytokines, chemical drugs/prodrugs or radioisotopes. Antibodies recognizing specific tumor cell surface markers are also valuable for diagnostic imaging. Examples of functional antibodies as cancer therapeutics include antibodies against EGFR and HER2, and cell surface antigens such as CD20, CD33, CD52, and PD1/PDL1. Despite the recent progress in drug target identification, a systematic effort to collect monoclonal antibodies that have functional blocking/stimulating or stimulating activities have not been initiated. The technology described here is to generate MAbs that exert physiological functions to living cells through membrane proteins. In the same process, novel membrane proteins and their functions are discovered.

Conclusions and perspectives

RbcMAb has been demonstrated to be superior reagent for many applications in lab including Western blots, immunohistochemistry and flow-cytometry. Due to its special antibody repertoire, high affinity and easiness for generation and genetic engineering, RbcMAbs are becoming an attractive source of anti-cancer therapeutics. A number of RbcMAbs will be used in targeted therapy against cancers in near future.