Industrially-produced antibodies are rarely found doing what antibodies evolved to do naturally, namely fighting disease. Instead, most are typically used as sensors, including in pregnancy test kits, or as the highly specific binders for enzyme linked immunosorbent assay (Elisa) and other detection applications.
But antibody therapeutics are gaining in importance, despite last year’s setback, when dangerous side-effects of an antibody product only became apparent during a clinical trial (C&I, 2006, 7, 8). A growing number have been approved for use in the clinic, mostly for various kinds of cancer, while Germany-based biotech major Morphosys alone has around 40 antibody products in its pipeline. In 2005, the sales of antibody therapeutics exceeded $13.5bn, and are expected to rise steeply as more of them become available to fight a wider range of diseases.
Despite this progress, the actual production of antibodies has remained difficult and expensive, involving mammalian cell cultures such as Chinese Hamster Ovary (CHO) cells. The requirement for such cell cultures has held the market price of monoclonal antibodies extremely high, with milligram quantities costing hundreds of dollars and ruling out many medical applications.
The hunt is on for alternative routes that might make therapeutic antibodies more accessible and affordable. All antibodies are composed of two immunoglobulin proteins, comprising both a heavy and a light chain of amino acids, respectively. While other cells can produce the amino acid chains of an antibody just as well as mammalian ones, it is the final ‘decorations’ of the molecule that cause worries in foreign environments. Specifically, there are crucial disulphide bonds stabilising the three-dimensional structures, and specific sugar groups need to be attached in specific places – the protein is glycosylated – for an antibody to function normally in a physiological environment.
While bacteria have proven useful for producing many proteins, they are not very good at expressing antibodies in large quantities. Plants have been made to express antibodies since the late 1980s, but have the drawback for therapeutic applications that they produce different sugar modifications, or glycosylation patterns, than mammalian cells. In 2005, researchers at the Centre for Genetic Engineering and Biotechnology (CIGB) in Havana, Cuba, nevertheless developed a plant expression system for the mouse monoclonal antibody against the hepatitis B virus surface antigen,¹ the key component of a successful vaccine against the disease. The antibody is crucial for purification of the antigen and has already obtained government approval for use in the Cuban pharmaceutical industry.
Express expression
But while transgenic plants can produce good yields of protein, the process from genetic modification to harvest takes months, or even years. Last summer, the group of Yuri Gleba at Icon Genetics in Halle, Germany, reported a novel, high speed plant expression system based on simultaneous infection with two non-competing viruses, to produce gram quantities of the precious proteins within two weeks.²
Specifically, the researchers infected tobacco plants, Nicotiana benthamiana, with tobacco mosaic virus (TMV) and potato virus X (PVX) carrying the genes for heavy and light chains of an immunoglobulin of the IgG class. Using this procedure, they produced around half a gram of functional antibody per kilogram of fresh leaves – a result CIGB’s Merardo Pujol describes as ‘an outstanding development’.
Antibody expert Andreas Plückthun from the University of Zurich, Switzerland, also welcomed the new method, saying that it ‘could potentially be of interest as an alternative production method, if it becomes economically attractive.’ However, he also warned that ‘the applicability of such plant-produced antibodies may be somewhat limited,’ because of the lack of mammalian-style glycosylation patterns.
‘IgGs produced in plants will normally not be able to bind to Fc receptors because the plant glycosylation usually prevents this, as does the unglycosylated form described by the authors,’ he says.
Fc or ‘fragment, constant’ receptors, found in various kinds of immune cells, bind to the unchangeable parts of antibodies once these have found their antigen, and thus connect the recognition of a target to its destruction.
Plückthun’s fears have been addressed by another group of researchers led by Kevin Cox at Biolex Therapeutics in Pittsboro, North Carolina. This company has developed a protein expression system, known as LEX, based on a fast growing water plant called lesser duckweed. With a doubling time of 36 hours, the lesser duckweed can be rapidly cultivated in a contained and controlled format like traditional cell culture expression systems, such as CHO, yeasts or bacteria.
Cox and his coworkers have now added a crucial ingredient to this approach, namely a gene silencer that suppresses the enzymes responsible for the plant-specific glycosylation patterns. They used the increasingly popular silencing method known as RNA interference (RNAi). Thus, in addition to the genes for the antibody heavy and light chains, they also equipped the plant with a new gene that produces a small RNA molecule designed to suppress the production of the enzymes that would normally attach fucose and xylose sugar groups to suitable glycosylation sites.
The researchers used this method to express the antibody MDX-060, which is directed against the CD30 antigen and is a potential treatment for two rare cancers: Hodgkin’s lymphoma and anaplastic large cell lymphoma. Producing this immunoglobulin in the duckweed system, the researchers found that the protein obtained is indistinguishable from MDX-060 expressed the traditional way in CHO cell cultures as far as in vitro tests are concerned. In some functional assays that affect the performance of an antibody in vivo, the researchers even claim that the plant-produced protein out-performs the one from mammalian cell culture.³ Cox claims the result is ‘a significant step forward in the production of recombinant glycoproteins in plants.’
Meanwhile, Gleba’s group has added another plant-based tool to the antibody production kit, by covering tobacco mosaic virus with protein A, a bacterial protein that binds antibodies and is commonly used in their purification.4 TMV can accumulate to around 1% of the leaf biomass, making its coat protein the single most enriched protein that can be harvested from plants. By genetically coupling this coat protein with protein A, Gleba’s team has created nanoparticles with a high affinity for immunoglobulins. As these particles could bind to many antibody molecules at once and can be easily separated from the solution by centrifugation, this approach could provide a surprisingly inexpensive way of purifying antibodies.
Green future
If and when more antibody therapeutics survive their clinical tests, plant expression systems could be promising candidates for their production. Other companies are also steering in similar direction. For example, Greenovation in Freiburg, Germany, is developing a bioreactor based on moss, with a transient expression system that is also geared at producing ‘humanised’ antibodies in the plant.
Companies that specialise in the contract manufacture of proteins are already looking forward to an increasing demand. Sigma-Aldrich subsidiary SAFC at St. Louis, US, has recently invested $16m in an expansion of its protein production facilities, which use both recombinant and natural plants.
How long will it take for the plant approach to bear fruit? Biolex’s Cox expects – clinical results permitting – to see a duckweed-produced monoclonal on the market by 2012. He also hopes that the increased functional activity will help to bring the cost of treatments down. Thus, for therapeutic antibodies, the future may be bright green.
References
1. Pujol M et al, Vaccine 2005, 23, 1833.
2. A Giritch et al, Proc. Natl. Acad. Sci. USA 2006, 103, 14701
3. Cox K M et al, Nature Biotech. 2006, doi 10.1038/nbt1260
4. Werner S et al, Proc. Natl. Acad. Sci. USA 2006, 103, 17678.
Features