For close to two decades, realization of the promise of monoclonal antibody (mAb) technology for the generation of therapeutic “magic bullets” has been challenged primarily by limited efficacy and safety related to immunogenicity of mouse antibodies in human patients. Among the technologies developed to overcome these hurdles were transgenic mice genetically engineered with a “humanized” humoral immune system. One such transgenic technology, the XenoMouse, has succeeded in recapitulating the human antibody response in mice by introducing nearly the entire human immunoglobulin (Ig) loci into the germline of mice with inactivated mouse antibody machinery. XenoMouse strains have been used to generate a large array of high-affinity, potent, fully human antibodies directed to targets in multiple disease indications, many of which are advancing in clinical development. Full validation of the technology has been achieved with the recent regulatory approval of panitumumab, a fully human antibody directed against epidermal growth factor receptor, for the treatment of advanced colorectal cancer. The successful development of panitumumab, as the first antibody derived from human antibody transgenic mice, signifies an important milestone for XenoMouse and other human antibody transgenic technologies and points to their potential contributions for future therapeutics.

RATIONALE FOR DEVELOPING HUMAN ANTIBODY-PRODUCING TRANSGENIC MICE

The discovery of hybridoma technology in 1975 for the isolation of high-specificity and high-affinity mouse monoclonal antibodies (mAbs) opened the door to a new class of therapeutics with a potential to substantially impact both therapy and diagnosis of many human diseases.

The Fc region of an antibody is the central link between the targeted antigen and the immune system. It is responsible for mediating a spectrum of effector functions that monoclonal antibodies (mAbs) use against tumors and pathogens. Whereas historically drug developers have kept the Fc region fixed, over the past decade there has been substantial effort to engineer it for improved effector function activity. This new direction has grown from a more mature understanding of the role of immune receptors in antibody therapy and the development of Fc modifications to control antibody/receptor interactions. In this chapter, we discuss how Fc engineering is being used to enhance antibody therapeutics for cellular effector functions, complement-mediated activities, and pharmacokinetic properties.

SITES FOR ENGINEERING AND OPTIMIZABLE PROPERTIES

The Fc region mediates binding of the antibody to all endogenous receptors other than target antigen. Although vaguely defined, an antibody's Fc region typically refers to the C-terminal portion of the hinge and the CH2 and CH3 domains, approximately residues 226 to the C-terminus using the EU numbering scheme. The human effector ligands that bind Fc can be divided into three groups (Figure 10.1): FcγRs, complement protein C1q, and the neonatal Fc receptor FcRn. The FcγRs all bind to essentially the same site on Fc, specifically the lower hinge and proximal CH2 region. Interaction with these receptors can elicit a variety of cellular effector functions that destroy target cells and regulate the immune system.

Since 1890, when von Behring and Kitasato reported that animal antitoxin serum could protect against lethal doses of toxins in humans, antisera have been used to neutralize pathogens in acute disease as well as in prophylaxis. Antisera are also used in vitro as diagnostic tools to establish and monitor disease. However, antisera invariably induce an immune response resulting in joint pains, fevers, and sometimes life-threatening anaphylactic shock. Various proteins contribute to the immunogenicity, as the serum is a crude extract containing not only the antibodies against the disease-causing antigen (often at low concentration), but also other antibodies and proteins.

FULLY MOUSE TO FULLY HUMAN

In 1975, Köhler and Milstein (1975) at the Medical Research Council's (MRC) Laboratory of Molecular Biology in Cambridge (UK) reported their discovery of a way to produce custom-built antibodies in vitro with relative ease. They fused rodent antibody-producing cells with immortal tumor cells (myelomas) from the bone marrow of mice to produce hybridomas. A hybridoma combines the cancer cell's ability to reproduce almost indefinitely with the immune cell's ability to produce antibodies. Once screened to isolate the hybridomas yielding antibodies of the required antigen specificity and affinity – and given the right nutrients – a hybridoma will grow and divide, mass-producing antibodies of a single type (monoclonals). Nearly a century before, the German scientist Paul Ehrlich envisaged that such entities could be used as magic bullets to target and destroy human diseases, and hybridomas seemed like a production line of batch consistency for these magic bullets.

The study of immunology is inexorably linked to the practice of animal husbandry. For example, the word “vaccinate” is derived from the Latin vaccinus meaning “of or from cows.” The name stems from the practice of protecting people from the deadly smallpox virus by inoculating them with an extract derived from sores of cow udders infected with the innocuous cowpox virus. Later, the serum of animals that had been repeatedly exposed to sublethal doses of diptheria toxin was shown to protect humans against diphtheria, a discovery that eventually led to the discovery of antibodies. Eventually the study of antibody-producing cells in mice led to the invention of monoclonal antibody technology by Kohler and Milstein in 1975. Thus, it is no surprise that germline engineering of the mouse was put to immunological use soon after this powerful technology was developed. Here I describe the VelocImmune® mouse created several years ago by megabase-scale humanization of the variable portion of mouse immunoglobulin (Ig) loci, by far the largest such precision genome-engineering project to date, and compare it with other methods for the generation of humanized or fully human monoclonal antibody therapeutics.

ANTIBODY THERAPEUTICS

Monoclonal antibodies have numerous advantages as drugs. They possess the qualities of (1) high affinity and exquisite specificity leading to few off-target effects and generally superb safety profiles, (2) long half-life leading to infrequent dosing, and (3) reproducible physical characteristics leading to routine production and shortened development time lines.

The discovery of monoclonal antibodies by Kohler and Milstein in 1975 sparked the generation of novel drugs that could be used to antagonize functional receptors of the immune system. The anti-CD3 antibody, OKT3, was the first of these drugs to be exploited clinically in the treatment of acute allograft rejection. Although the antibody was efficacious, neutralizing immunogenicity and, in particular, the often severe “flu-like” cytokine-release syndrome associated with initial doses of the antibody limited its application to other indications. As a consequence, the emergence of other immune-modulating CD3 or T cell-directed antibodies as therapeutics took a surprisingly long time. Three scientific developments rekindled interest in immune-modulating therapeutic antibodies resulting in many more antibody candidates entering clinical trials. The first development was the discovery that co-receptor CD4 antibodies could be used to tolerize to other proteins, thus establishing tolerance as a therapeutic paradigm. The second development was the discovery that rodent antibodies could be reengineered or reshaped to minimize their immunogenicity. Finally, the third development was the discovery that transplantation tolerance induced by co-receptor blockade was “dominant” and dependent on the induction of CD4+ regulatory T cells through so-called infectious tolerance. These findings together suggested that antibodies might be used sparingly to recruit the host's own tolerance mechanisms without evoking neutralizing responses.

Further studies in transplant models indicated that anti-CD4 therapeutic antibodies alone were insufficient when CD8+ T cells were also involved. In those circumstances, antagonism of CD8 function was also required.

The potential of antibodies as magic bullets for curing disease has excited the imagination of medical researchers ever since this phrase was first coined by Paul Ehrlich about a century ago. Seventy-five years after the publication of Ehrlich's side-chain theory to explain antibody-antigen reactions in 1900, Georges Köhler and César Milstein invented a means of cloning antibodies with defined specificity that paved the way for major advances in cell biological and clinical research. They were awarded the Nobel Prize in Medicine in 1984 for this ground-breaking research. In 1986, the first monoclonal antibody, the murine mAb OKT3 for preventing transplant rejection, was approved for clinical use, and although many other murine mAbs were subsequently investigated as therapeutic agents, most of them had a disappointing clinical profile largely due to their immunogenicity. This situation improved dramatically with the advent of techniques to humanize existing mAbs, followed by technologies that sought to imitate the generation of specific antibodies by the immune system in vitro. For example, the expression of antibody fragments in E. coli using bacterial leader sequences and the use of phage display and later ribosome display facilitated the selection of specific human antibodies from extremely large libraries. The process of somatic hypermutation to increase antibody affinity was mimicked by introducing random mutations. Another major advance for obtaining human antibodies was the creation of transgenic mice carrying a large part of the human antibody gene repertoire, which could be used to produce human antibodies by standard hybridoma technology.

Monoclonal antibodies have been used in a variety of ways in the management of cancer including diagnosis, monitoring, and treatment of disease. The U.S. Food and Drug Administration (FDA) has approved numerous monoclonals for the treatment of cancer (Table 13.1). Among the unmodified monoclonal antibodies, Panitumumab (Vectibix), cetuximab (Erbitux) and bevacizumab (Avastin) are now marketed for metastatic colorectal cancer, trastuzumab (Herceptin) for breast cancers that overexpress HER-2 receptors, and alemtuzumab (Campath) for B cell lymphocytic leukemia (B-CLL). Several other monoclonal antibodies are in late-stage clinical trials. With the general availability of these agents, it appears that antibody-based therapeutics have an established role in clinical oncology.

Radio-immunotherapy (RIT) utilizes an antibody labeled with a radionuclide to deliver cytotoxic radiation to a target cell. In cancer therapy, a monoclonal antibody (mAb) with specificity for a tumor-associated antigen is used to deliver a lethal dose of radiation to the tumor cells. The ability of the antibody to specifically bind to a tumor-associated antigen increases the dose delivered to the tumor cells while decreasing the dose to normal tissues. While antibodies armed with drug conjugates and immunotoxins kill only the targeted cell, radionuclide conjugates can exert a bystander effect, destroying adjacent cells that lack antigen expression. With external beam therapy, only a limited area of the body is irradiated. However, RIT, like cytotoxic chemotherapy, is a systemic treatment that, in principle, can eliminate metastatic disease throughout the body.

The majority of therapeutic monoclonal antibodies (mAbs) on the market and in development focus primarily on a limited set of targets selected on the basis of a few well-studied pathways. Truly novel targets (and their corresponding therapeutic mAbs) are rare and carry increased risk and challenges to develop because they, or the pathways they are involved in, are often neither well characterized nor extensively validated. The Raven therapeutic mAb discovery platform is especially efficient in discovering novel targets. Because the platform utilizes intact, living cells as the immunogen – and thus targets antigens present on the membrane of living cells – it is not biased upfront toward a particular protein, protein family, or signaling pathway. In addition, the presentation of these membrane targets in their fully processed and modified configuration and orientation in the living cell enables the discovery of mAbs to conformational epitopes as well as post-translationally modified epitopes. These epitopes may have greater tumor specificity and antitumor activity than those raised from less biologically relevant input such as purified or recombinant proteins and peptides. These epitopes can include binding sites on carbohydrates or lipids as well as conformational epitopes. In fact, the ability to discover these specific and active epitopes, not obvious when looking at mRNA or protein sequences, may open an entirely new class of antibody targets for cancer and other diseases. RAV12 is one example of a mAb that targets a carbohydrate epitope.

IgM antibodies exist both in a pentameric soluble form and as membrane-bound monomers mainly on the surface of naïve B cells, where they are part of the antigen receptor complex. Naïve B cells, constituting 75% of the peripheral blood B cell repertoire in humans (Klein et al., 1997), contain the largest diversity of an individual's rearranged immunoglobulin genes. The naturally occurring antibody repertoire contains specific antibodies against various antigens. In a primary immune response, B cells expressing antigen-specific IgM molecules are activated and differentiate into antibody-producing and -secreting plasma cells. Secreted antigen-specific IgM molecules are the first immunoglobulins occurring during a primary immune response. On the other hand, so-called natural antibodies exist independently of antigenic stimulation and are thought to contribute to the first line of defense against infections (Carsetti et al., 2004; Ochsenbein & Zinkernagel, 2000) as well as malignancy (Brändlein et al., 2003).

In addition to antibody-secreting plasma cells, memory B cells are generated during a primary immune response, a process that includes somatic hypermutation in the germinal centers. Most of the memory B cells have undergone a class switch and do not express IgM. In humans, however, IgM molecules with somatic mutations have been identified (Van Es et al., 1992). These somatically mutated IgM molecules contribute to an individual's immunological memory and constitute about 10% of the total peripheral blood B cell repertoire (Klein et al., 1997). IgM-expressing memory B cells protect against infections by encapsulated bacteria, and develop during the first year of life (Kruetzmann et al., 2003).

Nowadays, monoclonal antibodies are the fastest growing class of biopharmaceuticals. By the end of 2007, the U.S. Food and Drug Administration (FDA) had approved 21 therapeutic antibodies. Since 1975, with the seminal work of Köhler and Milstein (Köhler & Milstein,1975) describing the use of hybridoma technology for monoclonal antibody generation, major advances in the field allowed for the development of antibody libraries using recombinant technologies (reviewed by Hoogenboom, 2005, and Sergeeva et al., 2006). Various display technologies and the integration of automated screening methods now enable researchers to quickly identify multiple target specific antibodies for later development as biopharmaceuticals.

This chapter will look at MorphoSys's latest fully human antibody library, the Human Combinatorial Antibody Library HuCAL GOLD based on phage display of Fab antibody fragments. Besides a comprehensive introduction of the design and generation of the library, the chapter will describe the HuCAL-specific CysDisplay technology, explore the use of MorphoSys's proprietary AgX technology, and give some examples on the use of HuCAL-based antibody optimization by using standard affinity maturation approaches or the recently developed RapMAT technology.

HuCAL CONCEPT

The HuCAL technology is a unique and innovative concept for the in vitro generation of highly diverse fully human antibodies. The structural basis for the HuCAL libraries is provided by seven heavy chain and seven light chain variable region genes (Knappik et al., 2000).

In the language of modern biotechnology, monoclonal antibodies (Köhler & Milstein,1975) were the first “library” of proteins that was available, and the immune system was the first “selection” technology by which a specific binder could be obtained. However, only the subsequent introduction of molecular biology into this field allowed a true control over the molecules (reviewed, e.g., in Plückthun & Moroney, 2005; Weiner & Carter, 2003). This development of technologies was largely driven by the desire to use antibodies therapeutically, since the extraordinarily strong immune response to a nonhuman antibody in humans had put an end to essentially all of these endeavors. As will be illustrated in the following paragraphs, technological developments intended to solve this problem made not only the use of an animal immune system, but, ironically, also the antibody molecule itself dispensable.

Three fundamental approaches have been developed to arrive at antibody molecules that are able to evade the human immune surveillance and which, at least from this perspective, may become potential therapeutics. The first approach, termed “humanization” (Jones et al., 1986), converts an existing murine antibody obtained by immunization into an analogous one with as much human sequence as possible. Another approach, a technical tour de force, was to introduce human antibody genes into a mouse and inactivate or delete the murine loci, such that an immunized mouse would then produce antibodies after immunization that essentially consisted of human sequences (Fishwild et al., 1996; Mendez et al., 1997).