Identifying monoclonal antibodies with high therapeutic potential is not an end in itself. To qualify for clinical trials and potentially enter a coveted market, the industrial development of biomolecules uses its own rules and technologies. Pathologies with a high prevalence (oncology, inflammation, etc.) may require several hundred kilograms of antibodies per year.
Considerable investment in infrastructure must therefore be anticipated well in advance of any marketing authorization. In parallel, the choice of expression system, equipment and production platforms to be implemented are all critical considerations that the developer must face.
In a competitive and regulated context, the challenge is to reduce development times and costs, without any compensation for the quality requirements expected for the final product. We will review the main production systems used to date and their expected short-term developments.
Therapeutic antibodies are a well-accepted class of drugs, particularly in the field of oncology, inflammation and organ transplantation. To date, more than 30 antibodies or antibody fragments are registered and available on all or part of the world market.
These molecules represent in 2008 a turnover exceeding 20 billion dollars and the growth prospects are estimated at 14% per year, far ahead of most other therapeutic classes. This success should continue with a tripling of the number of monoclonal antibodies ( mAbs ) that have entered clinical phases in the last ten years. The mAbs combine two therapeutic advantages: their good physiological tolerance and extraordinary specificity for their target.
These benefits are however mitigated by the cost of treatments linked to the complexity of development and production of mAbs . In fact, it will take 10 to 12 years and approximately $ 1 billion to move from discovery to the commercialization of an ATM . Most commercial mAbs are produced from mammalian cell lines. These technologies require a complex production environment, highly specialized teams and expensive raw materials.
Why is the production so expensive ?
Today, the global production capacity of mAbs is limited if we consider the number of these molecules in development and their chance of success on the market. However, we will see that the improvement in cellular productivities moderates this limitation.
The construction of new commercial production sites is nevertheless a necessity. These multi-year projects require anticipation, risk-taking and a budget of at least $ 500 million. Even if new horizons open for therapeutic applications of mAbs , it is necessary to foresee the multiplication of molecules covering identical or similar indications, and consequently, a decrease in the market share of each mAb .
Reducing overall development and production costs has become a major issue. In this article, we will review the current methods of producing mAbs and describe the technological innovations in development and production that will most likely be economically advantageous in the next few years.
Cell lines: highly specialized factories
The antibodies are molecules which structure primary complex. They also have secondary and tertiary structural patterns which largely influence their activity. The synthesis of mAbs biologically active involves sophisticated biochemical mechanisms such as trafficking in different cellular compartments, post-translational modifications and finally secretion.
The production system and the physicochemical conditions allowing the synthesis and the purification of mAbs have a major impact on the yields and the quality of the molecule.
From the late 1970s until the development of recent cell engineering methods, only certain mammalian cells have been used for commercial production of mAbs . In the near future, alternative systems may be added to mammalian cells for the production of mAbs.
The very first therapeutic antibodies were produced using hybridoma technology developed in the 1970s by Köhler and Milste.
The discovery was major and earned them the Nobel Prize in 1984. The B lymphocytes ( splenocytes ) of a mouse immunized with the antigen of interest are fused with immortalized cells that do not produce antibodies. The in vitro culture of this cell mixture is carried out in the presence of an inhibitor system allowing only the survival of fused cells or hybridomas.
Clones (from a single hybridoma producing a single type of antibody) can be isolated and used for large-scale antibody production. Although the first recombinant antibodies were produced with this technology, it has serious drawbacks.
The mAbs are of murine origin and their repeated injection into humans triggers an immune reaction with the formation of antibodies directed against these mouse antibodies. This defense mechanism results in the rapid elimination of murine antibodies from the body, thus seriously reducing therapeutic effectiveness.
About 80% of patients who received the first commercial murine antibody, OKT3, used to prevent transplant rejection, have been shown to develop an immune response. This immunogenicity of murine antibodies in humans explains that there are currently only five mouse mAbs on the market.
The lined cell of mammals
Murine hybridomas with human antibodies
Advances in molecular biology have overcome the limitations of hybridomas. After selection of the hybridoma of interest, the DNA sequences coding for the antibody are isolated. Gradual approaches have been developed to give the mouse antibody an increasingly human appearance.
The chimeric antibodies are constructed from the constant regions of human immunoglobulins to which are attached the variable regions of the antibody produced by the corresponding hybridoma. Subsequently, immunization techniques have made it possible to construct antibodies in which only the hypervariable regions of the mouse antibody are grafted in place of the equivalent sequences of a human antibody.
Ultimately, production of fully human mAbs became possible with the use of transgenic mice or phage libraries that combine the diversity of human immunoglobulin genes. The mAbs Vectibix ®, Humira ® and Stelara ® were obtained by these two has PPROACHES and are 100% human .
After all these manipulations which, it should be remembered, take place at the level of the gene, the DNA sequence obtained is cloned into an expression vector and the latter is introduced into a mammalian cell by transfection. Integration of foreign DNA into the cell’s chromosome is a prerequisite for future expression of the mAb . However, this mechanism occurs with a relatively low frequency.
The cells having integrated the vector are generally identified and isolated using a selection marker co-located on the expression vector and which codes for a sequence conferring on the cells the capacity to proliferate in culture in a selective medium.
Production of humanized mAbs in cellular systems
Many immortalized cell lines could potentially be used for the production of mAbs . In fact, with the exception of the two mAbs produced from hybridomas, all the commercial antibodies are produced in mammalian mouse (NS0 and SP2 / 0 lines, two murine myeloma lines) or Chinese hamster lines (CHO line derived ovary).
The latter is now favored by industrial researchers since it is used to produce around 50% of commercial mAbs and mAbs in clinical trials.
These three lines can be grown in suspension and are fairly well adapted to the change of scale ( scale- up) of cultures, from the in vitro stage in the laboratory to cultures in large capacity bioreactor (20,000 l). Cellular productivity is one of the essential criteria from an economic point of view. Spectacular progress has been made and productivity has increased twentyfold over the past ten years.
Currently, industrial production is of the order of 1 to 4 g / l of Acm and we will see below, which can be considered to exceed these productivity thresholds. The glycosylation profile of IgG produced from NS0, SP2 / 0 or CHO lines generally presents subtle variations which can nevertheless have a significant impact on the immunogenicity of the molecule.
These glycoforms are aujourd ‘ hui well be characterized and quantified. In addition, the mastery of production processes constantly evaluated by rigorous controls ensures consistency in the rate of glycoforms and an excellent safety profile in the clinic.
As we have seen, the cell lines used come from rodents and the viruses they integrate have been a source of concern. Over the past fifteen years, in-depth studies have been carried out to ensure the absence of human pathogenic viruses in NS0, SP2 / 0 or CHO cells. Nevertheless, specific measures are included in the purification processes to remove the viral particles before the use of mAbs in the clinic.
Emerging antibody production systems
New cellular models
New cellular models could soon upset the established order with the rodent lines. The Per.C6® line comes from immortalized human retina cells and has a very high density culture potential. Several studies have demonstrated the capacity of Per.C6® to produce a mAb with yields reaching 10 g / l. By nature, recombinant proteins from this line have a human glycosylation profile.
These advantages are decisive elements to allow the Per.C6® line to become a production platform for Acm . Other lines are currently being used in development: avian cells EBx ® from the company V ivalis and rat myeloma YB2 / 0 from Laboratoires LFB (French laboratory for fractionation and biotechnology).
Their main advantage is to produce molecules with a low rate of fucose , which in principle makes it possible to increase their capacity to trigger the cytotoxicity dependent on antibodies (ADCC), one of the effector functions of mAbs . Bacteria, yeasts, filamentous fungi, microalgae , transgenic plants and animals are all production systems that can potentially be used for the production of antibodies .
The yeast Pichia pastoris have modified are genetically engineered to allow the expression of proteins with human glycosylation patterns.
The immunogenicity of the recombinant biomolecules is thus greatly reduced. Although the productivity of Pichia pastoris is significantly lower than that of the mammalian cells described above, recent publications describe protein titers exceeding 2 g / l .
This economically advantageous expression system could therefore be used in the near future for industrial productions. Unlike the rodent lines mentioned above, all of these emerging systems that are not yet registered will have to face meticulous regulatory expertise in order to qualify as a production system.
Improvement of cell lines by genetic and metabolic engineering
The insertion of the expression vector at the chromosome level is a random phenomenon. If integration occurs in a highly condensed region of the chromatin, a low level of expression of the recombinant protein will generally be observed.
The genomic environment of the integration site and the presence of epigenetic modifications of DNA (methylation) and histones (acetylation, methylation, phosphorylation, etc.) considerably influence the transcription efficiency of the transgene. The STAR ( stabilizing and antirepressor ), S / MAR (matrix associated regions1) or UCOE ( ubiquitous chromatin opening elements2) approaches, to name a few, allow the addition of specific DNA sequences near the coding insert. the mAb and induce chromatin environment favorable to the transcript .
These new technologies can potentially simplify the selection of cells with a high production yield after transfection and also reduce development times. Metabolic engineering of cells also makes it possible to improve or control the quality of the mAbs produced. The presence of specific glycosylation patterns influences the interactions of mAbs with certain cells of the immune system.
Thus, a CHO line from which the fucosyl transferase gene has been deleted has enabled the expression of non- fucosylated Acs which trigger ADCC activity multiplied by 100. On the other hand, modifications of the Fc region of the Acm have have been introduced which increase the affinity of this region towards its receptor, improving the stability of the antibody in vivo. Cellular modifications have also been described to prolong the duration of cell culture in a bioreactor.
One of the strategies is based, for example, on the use of interfering RNAs to inhibit proapoptotic genes such as Bax , Bak or caspase 3. Conversely, additional copies of anti-apoptotic genes such as Bcl-2 and Bcl – XL were inserted into the producer cell, increasing
its survival, and therefore resulting in increased production yields. In the case of a line overexpressing the Bcl -XL protein , an 80% increase in the volumetric productivity of a humanized antibody has been demonstrated by comparison with that of the parental cell.
The improvement in transcriptional activity induced by these modifications can have consequences on the cellular capacity for assembly of polypeptides such as antibodies. If they are incorrectly or incompletely folded, they will be secreted in less quantity and their structure will be heterogeneous. At this stage, chaperone proteins play an essential role in ensuring the assembly, cell traffic and secretion of antibodies.
Several works describe how it is possible to modify this cellular machinery, by playing for example on the relative levels of expressionof BiP proteins ( immunoglobulin heavy chain binding protein ) and PDI ( protein disulfide isomerase ).
Many other approaches have been documented, the ambition of which is to improve the potential of cell lines for the production of recombinant antibodies (see the review by Barnes). Even if these general principles are for many at the experimental stage, it is a safe bet that certain approaches will develop until being applied at the industrial stage in the coming years.