Technologies providing access to antibodies as a class of drugs are sought after in the industry. Dr Achim Knappik, head of r&d at AbD Serotec, a division of MorphoSys, looks at their origins and recent developments
Antibody technology has developed significantly since Koehler & Milstein used cell fusion to produce monoclonal antibodies from immunised mice back in 1975. Therapeutic antibodies made their debut in 1986 when Johnson & Johnson's mouse-based antibody OKT3 was approved to fight transplant rejection.
However antibodies derived from mice soon proved to be of limited use as therapeutic agents, as the human immune system recognises such antibodies as foreign molecules and may trigger a defence reaction against them.
Since then therapeutic antibodies have proceeded from murine origin to chimeric and humanised antibodies, and decreased their side effects slightly with every step they took.
More recent technological advancements have introduced fully human antibodies, which should show very few or no side effects at all. The antibody Humira became the first fully human antibody therapeutic to gain market approval in 2002. Today, only a small and shrinking number of companies worldwide can provide access to fully human antibodies for therapeutic purposes.
MorphoSys, founded in 1992, developed a concept for the in vitro generation of highly specific and fully human antibodies and established an antibody library called HuCAL (Human Combinatorial Antibody Library), which comprises in its current version HuCAL GOLD more than 10 billion different antibody specificities.
During the initial design of HuCAL, MorphoSys" scientists conducted a sequence and structural analysis of the human antibody repertoire addressing the following questions: how many variable domain frameworks are necessary to cover most of the different structures seen in the repertoire, and which amino acid is represented at which position of the antibody hypervariable regions and in how many cases?
Despite the great variability in parts of human antibodies, the analysis confirmed that the two main structural components, the so-called heavy and light chain, of human antibodies assume only a limited number of basic shapes or frameworks..
The stem of the Y-shaped immunoglobulin carrying most of the constant regions determines the antibody's isotype and hence certain functional properties of the antibody. Its sequence is very constant and varies only by the given isotype, so it could be disregarded in relation to the initial library. The specificity of an antibody is determined by the variable regions, and within these regions by the so-called complementarity-determining regions or CDRs.
At the protein level these six highly variable hot spots. form loops at the outer tips of the arms of the antibody structure, which together create the antigen-binding site. The correct conformation of the antigen-binding site is assured by the, so-called, variable framework regions. These shape-giving framework regions, the CDRs and the first constant domain of the heavy and the light chain, respectively, together form the so-called Fab fragment, which is basically the isolated arm of an antibody.
By assorting the antibody germ line sequences according to sequence homology and weighting their importance according to usage in the human immune response, seven master gene segments for both heavy chain and light chain variable framework regions could be identified. With this analysis concluded, the general structure of the planned human antibody library was set: seven heavy chain and seven light chain variable region genes form the basis of the HuCAL libraries and reproduce the structural diversity of the human antibody repertoire.
The combination of these genes gives rise to a total of 49 framework combinations in the master library. By superimposing the highly variable genetic CDR cassettes on these frameworks, the entire human antibody repertoire can be mimicked.
MorphoSys used trinucleotides as building blocks for cassette synthesis, thereby gaining full control over diversity at each position of the antigen binding site. This ensured a high percentage of functional antibody sequences in the initial library. Such a system proved to be able to recreate the complex interplay between CDRs and framework regions to encompass stability and variability in equal measure.
While the genetic construction of the library is its major feature, the way the antibodies are stored and most of all made available for selection is also a critical component. The MorphoSys antibody library is based on a proprietary phage display technology. This approach exploits the biological characteristic of bacteriophages viruses that infect bacteria while being completely harmless for humans.
The bacteriophages used in phage display libraries incorporate the genetic information for an antibody within their structure during assembly, and display this antibody protein on their surface. Antibodies displayed this way maintain their natural binding characteristics and structure, and can recognise a corresponding antigen. This principle called phenotype-genotype-coupling forms the basis of antibody selection in phage display libraries.
To give a brief example of a typical selection process at MorphoSys we'll accompany a target molecule or antigen through the different steps of antibody isolation and optimisation. The antigen X, which is for example indicated to play a critical role in a certain disease, is immobilised on an ordinary microtiter plate. Our library, displayed on bacteriophage particles, is then passed over that target and after a certain number of washing steps only those phages with antibody fragments recognising the target molecules remain bound.
As mentioned, the isolated bacteriophages also contain the genetic information for the antibody fragment. Therefore, when bound phages are amplified after elution from the antigen, the genetic information for the given antibody is also amplified. Amplification takes place by infecting E. coli with the eluted phages, which leads to massive production of daughter phages.
While antibodies obtained after such selection are suitable for use in the majority of applications, their use in therapeutic applications may require higher affinities. Once the most promising antibody candidates from the first selection are isolated, new libraries based on the candidate genes that contain amino acid changes in one of the six CDRs are created. These new sub-libraries with some 109 new antibody specificities can then be selected again against the target molecule using more stringent binding conditions than before. This technique allows isolating those new antibody variants that are capable of binding even better to the given target than the parental antibody.
One of the strengths of the HuCAL technology is the fact that these sub-libraries for antibody optimisation can be created fast and routinely on the genetic level. By having introduced unique restriction sites flanking the CDRs in all the HuCAL master genes, the most relevant structures for antibody affinity and specificity can easily be manipulated in a simple cloning step. This modular gene structure allows routine antibody maturation by molecular evolution principles (mutation and selection). Furthermore, the different pre-built CDR cassettes are diversified according to the natural repertoire of CDR sequences identified in the original bioinformatical analysis.
Trinucleotide-directed-mutagenesis or TRIM, a technology exclusively licensed to MorphoSys from the Johns Hopkins University, has proved a key tool in the maturation process as well as in the establishment of the initial HuCAL libraries itself. Using pre-assembled trinucleotides, TRIM allows the synthesis of any desired mixture of amino acids at will at every single position of the variable regions. This ensures complete control over the composition of the CDR regions to keep them consistent with natural human antibody sequences.
After several rounds of maturation HuCAL delivers Fab fragments, capable of binding a given target with very high affinities. If other formats than Fab fragments are needed, the modular gene structure of HuCAL allows easy switching between different antibody formats. Fab fragments can be converted via one cloning step into a dimeric format, which we call mini-antibodies ideally suited for application like Immunohistochemistry or Western blotting. On the other hand, when inserted into a standard immunoglobulin expression vector, fully human IgG antibodies can be produced ready for assessment as therapeutics. HuCAL therefore can provide a seamless transition from a rapidly generated antibody fragment to a fully human antibody drug candidate.
Short timelines in drug discovery and development put significant pressure on technology providers to further enhance their systems. MorphoSys is constantly upgrading its technology platform and has presented its RapMAT technology as a first answer to this trend in 2006.
RapMAT represents an in-built affinity maturation process. The basis for this technology is the modular concept of MorphoSys's HuCAL technology. Using RapMAT, the uncharacterised polyclonal output after two rounds of standard HuCAL selection is used as genetic basis, and new diversity is introduced by insertion of a pre-built CDR cassette library. This is in contrast to HuCAL's standard maturation process, where individual antibody candidates are selected and matured by subsequent CDR exchange. Subsequently, two further selection rounds are applied under high stringency conditions to select for high affinity. This ultimately leads to the direct selection of antibodies that have shown up to a 40-fold increased affinity for their target molecule. All resulting antibodies retain a fully human composition.
Providing therapeutic candidates for the pharmaceutical and biotechnology industry as well as for MorphoSys's in-house drug development programmes is currently the main application of the HuCAL technology. Almost half of the currently 45 active therapeutic antibody programmes are intended to cure various forms of cancer. Small inroads have been made in the in vitro and also in vivo diagnostics markets, both through partnerships but also through the activities of MorphoSys's second business unit AbD Serotec.
Within AbD Serotec, in the large part a traditional research antibody business, HuCAL is additionally used to provide custom generated antibodies against new target molecules for research customers. Until recently, all research antibodies were developed using animal-based technologies a method which has proved less than ideal. MorphoSys is confident that the market as a whole is ready for a technological shift and that in the mid- to long-term, animal-based methods will be replaced by in vitro approaches such as HuCAL GOLD technology.