From Concept to Diagnosis

Novocastra™ Antibodies

Kenneth Mitchell, Ph.D., Leica Microsystems

Newcastle upon Tyne – this lively town in Northern England is not only famous
for its culture, but it has been the base for the developer and manufacturer of
antibodies and other reagents of the brand, Novocastra™. “Novocastra” is based upon the Latin name for the city of Newcastle and was the name given to the original company founded by Prof C. Wilson Horne. The company grew and in 2003 was integrated into Vision Biosystems where its products have continued to make an impact in the immunohistochemistry diagnostic market.

How do you choose which antibody to make next?

Rees: For our selection process, obtaining the “Voice of our Customer (VOC)” is crucial. By conducting a VOC exercise we are able to obtain valuable input from the market, forge collaborative links with pathologists and obtain the data necessary for reagent design and development. We rely upon data within the current scientific literature and also upon feedback from our customers, through marketing and our network of Key Opinion Leaders (KOL ) in Pathology.

So how are they designed?

Rees: Antigen design and the selection of an appropriate target region is probably the most important step in antibody development. This determines the level of antigen solubility, immunogenicity and specificity of the monoclonal antibody produced. When developing an epitope-specific antibody, immunising proteins can either be made using synthetic chemical processes or produced via recombinant protein technologies using the encoding genes. Our scientists tend to favour the latter approach, producing recombinant protein and including as large a target region as possible. This maximises the probability of producing a robust monoclonal antibody suitable for use on FFPE tissue.

The antigen design process involves several important steps (Fig. 1). The first step in the design process is the homology analysis where we assess the target for regions of similarity to other proteins. We identify either the region to avoid or the region(s) to produce screening antigens to help select for those hybridomas producing antibody specific for our target. Once a decision on the target region has been made, we then design DNA primers specific to the encoding gene sequence to facilitate the amplification of the gene. The amplified genes are then transferred to a bacterial protein expression system where the protein is expressed and purified by means of column chromatography to produce soluble immunogen.

For any given target, we can produce a number of proteins that can either be used as an immunogen or a screening protein in ELISA and Western blot validation for any antibody produced. To maximise our success rate, we have developed a series of antigenic fusion tags to enhance immunogenicity of our antigens.

What are the next steps to develop an antibody for
commercial diagnostic utility?

Piggott: Firstly we have to identify that the mouse has responded to the immunogen by testing a sample of the mouse serum by IHC on FFPE tissue
and by Western blot. A positive result is indicated when the mouse has responded to the protein of interest by giving the correct staining pattern on tissue and when it recognises a protein of the correct molecular weight by blot. Five days prior to fusion, we would give that mouse an intravenous boost of antigen and a somatic cell fusion is carried out similar to that first described by Köhler and Milstein (1975). This process involves the fusion of antibody-secreting murine splenocytes to murine myeloma cells to give a hybridoma cell with immortal properties and antibody-generating ability. By screening the many hybridoma cells generated (Fig. 2), we are able to select for the best antibody for use in FFPE by immunohistochemistry. The next stage involves the stabilisation of the antibodysecreting properties of the hybridoma by growth in cell culture at which time we select hybridomas that grow well and continue to secrete the antibody of interest. Following this stabilisation, we screen the antibody produced on a range of normal and abnormal FFPE tissues obtained from ethically approved sources.

How do you ensure the quality of the reagents?

Piggott: All our research, development and manufacturing procedures comply with regional regulatory requirements. We perform a wide range of testing on the antibody to determine that it is the best performing, most sensitive and specific antibody that can be used both as a liquid concentrated antibody on the bench in manual staining and as a ready-to-use reagent for the Bond-max™ automated immunohistostainer. We have responded to the market by choosing an antibody that works in both settings. We also perform extensive stability experiments to determine that the product is stable as a liquid reagent and we produce multiple batches for manufacture and test each of these. Following this “in-house” assessment, we work closely with a number of independent pathologists who “road test” our antibodies on surgical pathology cases, alongside a competitor antibody, if one is available. It is through such extensive screening on a wide range of tissues that have been prepared in numerous institutions that we are able to ensure only the best possible antibody, or clone, is selected for release as a Novocastra™ reagent. Coinciding with this antibody release, our collaborating KOL ’s may write a peer-reviewed article for publication. One recent success story has been the development of an antibody to the protein, DOG-1, over-expressed in a high percentage of gastrointestinal tumours or GIST s as described and evaluated by Professor Marco Novelli for the Journal of Histopathology (in press).

How long does this process take?

Piggott: The time taken variable depends upon a number of different factors. If we obtain the correct, successful response in the mouse, the project usually takes between 18–24 months to complete. If the project is very important we may have several strategies running in parallel. The longest we have worked on a project was seven years to produce the world’s first CD33 that works in FFPE and that was described in the last edition of resolution. That project employed several serial strategies.