Monoclonal Antibodies in Therapeutics – a primer

Let me start with a quick re-cap of how Monoclonal Antibodies (MAbs) are made in the laboratory.

The journey of MAb therapeutics to its current state has been interesting, with several important iterations.

Murine: These were the first generation MAbs, in which immunized mouse splenocytes (spleen cells, including B-lymphocytes) were fused with a myeloma partner to form a hybridoma. The successful hybridomas would be screened for clones that would recognize the specific antigen. The resultant monoclonal antibody would have both heavy and light chains of mouse origin.

The first ever therapeutic MAb, described in 1986 and licensed by the US FDA in 1992, was a murine monoclonal against CD3 surface protein of T-lymphocytes; named ‘Muromonab-CD3’, this MAb of the class IgG2a was indicated for use to prevent heart, liver or kidney transplant rejection. In the World Health Organization’s International Nonproprietary Names (INN) system, a murine antibody is identified by the pre-stem -o- in its INN; muromonab, around since before INN existed, is the only exception. One example is edrecolomab, approved in Germany since 1995), a mouse IgG2a against a protein antigen called EpCAM (Epithelial Cell Adhesion Molecule), carried on the surface of many tumor cells.

Theoretically, a related species of MAbs may come from rats, one in which both chain types are of rat origin, and a rat antibody would be identified by the pre-stem -a- in its INN. However, currently there are no purely rat therapeutic MAbs in the market. There is a rat-mouse hybrid MAb, engineered with binding sites for two different antigens (class known as ‘trifunctional antibodies’) with the substem -axo- in its INN, called catumaxomab, but more about that later; by design, this MAb qualifies as the next-generation of therapeutic MAbs, the ‘Chimeric’ antibodies.

Murine MAbs, although in use, are considered to have limited therapeutic value. Recognized by the human body as foreign substances, the murine MAbs often elicit a human immune response (known as HAMA or “human anti-mouse antibody” response) which neutralizes the foreign antibodies, prevents their ability to engage immune cells as intended, and accelerates their removal. This, obviously, limits the effectiveness of murine MAbs, and may even result in deleterious side effects reminiscent of the ‘serum sickness‘ observed during the days of serum therapy, such as fever, chills, arthralgia and life-threatening anaphylaxis. The Chimeric MAbs came off the efforts to solve this problem.

Chimeric: These were the second-generation MAbs; a chimeric antibody is encoded by genes from more than one species, technically a fusion molecule, produced by genetic engineering methods. Remember the basic antibody structure from the FIRST PART of this series? For the heavy and the light chains, a chimeric chain contains a foreign Variable (V_H and V_L) domain (originating from one species other than human, or synthetic) which is connected to a Constant region (C_H and C_L) of human origin. The heavy chain constant region comprising the Fc portion is recognized by the immune effector cells for binding and bringing about many of the antibody’s cell-mediated effects, such as opsonophagocytosis, ADCC and so forth. Therefore, in a chimeric MAb, while the foreign variable regions confers epitope specificity for the intended therapeutic target, the human constant domain – introduced by genetic engineering to replace the analogous foreign constant domain – eliminates most of the potential immunogenicity of (i.e. the possibility of HAMA response against) the MAb without altering its specificity. Therapeutic chimeric antibodies are identified by the pre-stem -xi- in its INN; for example, abciximab, infliximab and the widely successful rituximab. There are other examples in the SECOND PART of this series.

The above mentioned Catumaxomab, which I considered ‘chimeric’, is a different type of hybrid; one half of the antibody (one heavy IgG2a chain and one kappa light chain) is of mouse origin, recognizing tumor-associated EpCAM, whereas the other half (one heavy IgG2b chain and one lambda light chain) is of rat origin, recognizing CD3 on the surface of T-lymphocytes. The hybrid Fc segment of this engineered antibody is capable of binding to immune effector cells like any antibody. Therefore, this MAb binds to a tumor cell via one arm, to a T lymphocyte via the other arm, and to an Fc-receptor bearing immune effector cells, such as macrophage, a natural killer cell or a dendritic cell, via its Fc portion, triggering an immunological reaction against the cancer cell. It has been approved for malignant ascites in the European Union, and currently undergoing phase II/III clinical trials in the US for gastic and ovarian cancers.

Some chimeric MAb may still have issues. For instance, cetuximab, approved by the US FDA for metastatic colorectal cancer, is known to give rise to HAMA IgEs.

Although beyond the scope of this description, here is a quick reference to other non-antibody chimeric proteins in therapeutic use. Etanercept, for example, is a TNFα blocker biologic drug, to make which the gene for human tumor necrosis factor receptor (TNFR) was recombined via genetic engineering with the gene for human IgG1 Fc segment. In the resultant chimeric protein, the TNFR part provides specificity for the target, and the antibody Fc segment ensures stability and localized drug delivery. Etanercept has US FDA approval to treat rheumatoid arthritis, psoriatic arthritis, plaque psoriasis and ankylosing spondylitis.

Humanized: These were the third generation therapeutic MAbs, designed to be an improvement upon Chimeric MAbs. In the “humanization” process, the antibody protein sequence is genetically modified to increase its similarity with human antibodies, by interspersing human sequences within the murine complementarity determining region (CDR) segments in the epitope-specific Variable domain; the goal is to further reduce the change of an unwanted HAMA response. A humanized therapeutic MAb is identified by the pre-stem -zu- in its INN; for example, trastuzumab and alemtuzumab. Other examples are in the SECOND PART. Nevertheless, induced antibody responses against humanized therapeutic mAbs (termed ‘HAHA’, human anti-human antibody response) have also been observed, giving rise to the concern that other factors, such as the nature of the antigen, the disease process treated and the schedule of administration, may contribute to the unwanted immunogenicity of the MAb.

Human or ‘Fully Human’: This represents the fourth generation of therapeutic MAbs. A human antibody is one of which both chain types, and the J chain in the case of polymeric antibodies, are of human origin. A human antibody is identified by the pre-stem -u- in its INN; for example, adalimumab and ustekinumab. Other examples are in the SECOND PART. Different technologies were devised and studied to aid in production of human MAbs.

• Phage Display: ‘Phage’ or filamentous bacteriophage is a virus that can infect and replicate within bacteria. It offers an elegant and incredibly powerful technique to produce antibodies in vitro. In short, the genes for the Variable (V) regions of antibody heavy and light chains can be inserted separately into these phages, and they display the proteins on the surface. These genes can come from (a) naïve B-cells, or (b) immunized, antigen-specific memory B-cells. The proteins may be selected by screening against the antigen of choice, and the desirable proteins may be produced and secreted by bacteria infected by the specific phage. The secreted proteins can be reassembled in vitro to make functional antibodies. This technology is still developing and has great potential, but it is cumbersome, requiring multiple different steps. The antibodies made with this technology may also suffer from random heavy and light chain pairing, reducing the antibody efficacy. An example of fully human phage display derived therapeutic MAb is adalimumab.
• Transgenic Mice: In early 1990s, transgenic mouse strains, in which mouse Ig genes were replaced by human Ig genes, were developed. This mouse strain contains human genes and produces human antibodies, and murine-murine hybridomas – which immortalize antigen-activated B-cells – can produce large quantities of antigen-specific human MAbs; examples are golimumab and ustekinumab. However, the antibody diversity (in epitope recognition) is created in a mouse strain environment, where the antibodies undergo affinity maturation, as opposed to a in human environment. Although this gave rise to the concern of limited germline repertoire engagement, now there are other mouse strains with larger human Ig transgenes to address that issue. Moreover, the results of clinical trials have now started to come in and should permit better analysis of the efficacy of these transgenic MAbs.
• Primarily human system:This, of course, is the best-scenario for MAb therapeutics – the generation of human MAbs in a human system, thereby eliminating unwanted immunogenicity of the therapeutic agents. However, for human MAbs, the generation of human hybridoma cell lines of acceptable stability has posed a particularly galling challenge. For one, antigen-primed human B-lymphocytes needed for fusion are difficult to obtain on a routine basis; therefore, several alternative strategies have been devised for that purpose – which, nonetheless, has not been without problems.
  1. Mouse myeloma cell line transfected with human immunoglobulin genes – secrete low levels of human antibody.
  2. Fusion of human B cells with a murine myeloma partner – barring a few exceptions, mouse-human hybridomas cease human immunoglobulin production due to disturbances in gene expression; also, the peculiar phenomenon of selective loss of human chromosomes is well known; for some reason, mouse-human hybrids seem to preferentially retain human chromosomes 14 (H chain) and 22 (λ-L chain), and lose chromosome 2 (κ-L chain).
  3. Fusion of human B cells with a mouse-human hybrid ‘heteromyeloma’ – designed to solve the problem of chromosomal loss, but are unstable and secrete low levels of antibody.
  4. Transformation (and consequently, immortalization) of antigen-primed human B-cells with lymphotrophic (i.e. preferentially infecting lymphocytes) Epstein-Barr virus (EBV, a.k.a. human γ-herpesvirus 4) – moderately successful technique, but these transformed cells show unstable growth and secrete low levels of antibody, ceasing production altogether after a variable time period.
  5. Fusion of human B cells with a human lymphoblastoid cell line (LCL) – unfortunately, such hybridomas tend to secrete antibody molecules derived via permutation of the antibodies from both the fusion partners.
  6. Fusion of an EBV-transformed human B-cell line with a murine or human myeloma or a human LCL – secretes large amount of antibody. Because of the chromosomal loss in murine-human hybrids, human-human hybrids are considered more preferable. EBV transformation is beneficial because it causes polyclonal activation of B-cells, generating a greater V-region diversity, and makes the B-cell more competent for hybridization. These hybrids are stable and more efficient, and can be made to secrete specific antibodies.

In the next post, I shall focus on some very recent developments in antibody therapeutics and their applications. Stay tuned!

This is the THIRD PART of a multipart series on exciting new developments in the world of antibodies.

ACKNOWLEDGEMENTS with gratitude: The images of the mouse and rabbit have been modified from the product listing page of Charles River laboratories; the images of the myeloma cell and hybridoma cell have been modified and adapted from Figure 5 of Karpas et al. (PNAS USA 2001, 98:1799).