Novel Platforms for Manufacturing Biosimilars: A Commentary

Biosimilars are cost-effective alternatives to innovator biologics used for the treatment of chronic debilitating and often life-threatening disorders, such as different types of cancers (breast cancer, non-Hodgkins lymphoma, metastatic gastric cancer, non-small-cell lung carcinoma), inflammatory diseases (ankylosing spondylitis, Crohn’s disease, psoriasis, rheumatoid arthritis, ulcerative colitis) and hormone deficiencies (diabetes, anaemia, pituitary dwarfism, etc).  However, the difference in cost between a biosimilar and its reference is only about 20 to 30%.  This contrasts sharply with the cost difference between a generic and its innovator, which can be as much as 90%.  One of the reasons is the stringent regulatory requirements for the registration of biosimilars compared to those for generics. It is estimated to cost between a 100 and 200 million dollars to develop a biosimilar, whereas a generic would cost only between 1 and 5 million dollars (1). 

Another reason for the higher cost of biosimilars is their manufacture and quality control.  The cost of infrastructure, including equipment, for production of a biosimilar is far in excess of that for a generic. Since it is unlikely that regulatory requirements for biosimilar approvals will change and, hence, that the cost of development will decrease, many start-up companies, including South African start-ups, have looked at alternative ways of manufacturing biosimilars.  One approach is to use plants, especially tobacco plant species (e.g., Nicotiana benthamiana), which have been used to express many mammalian proteins, including monoclonal antibodies (2).  The process is commonly referred to as biopharming and there are two ways to genetically modify plants to turn them into mini-factories or bioreactors for biologic drug production. One process is known as transient expression, which involves infiltrating leaf material with an Agrobacterium suspension (immersing the plants in the suspension under vacuum pressure) containing the genetic material for a target therapeutic protein or antibody.  The other method is to develop stable transgenic plants by altering the DNA of a plant’s nuclear or chloroplast genomes.  Seed lines are then developed for continual propagation of plant biomass, much like the development of cell banks for mammalian cell culture.  This method is, however, less popular than the former, mainly because of the length of time it takes to generate and select for the desired seed lines and also because of the much lower yield of recombinant proteins obtained compare to the transient expression system (3).

Biopharming offers many advantages over mammalian systems for biopharmaceutical production, including lower upstream costs, higher yields, speed of manufacturing, indefinite scalability and ease of handling.  Plants also offer the advantage of biological safety as there is no health risk from contamination with animal viruses (2).  However, using plants for expression of mammalian proteins also have limitations. One common argument against the use of plants for this purpose is their inability to perform authentic N-glycosylation. A major concern is the presence of beta 1,2-xylose and core alpha 1,3-fucose residues on complex N-glycans as these non-mammalian N-glycan residues may provoke immunogenicity reactions (4).  Today, even this challenge has been overcome by genetically engineering plants not to express these plant sugars.  Thus, it would appear that plants do offer an opportunity to reduce the cost of biosimilars even more.  Viewed from a regulatory perspective, however, would plants be an acceptable platform for the expression of an antibody product, such as trastuzumab or rituximab, which were originally developed with a mammalian cell expression system such as Chinese hamster ovary (CHO) cells? In my opinion, this would be difficult for the following reasons:

  • The CHO cell system, although it may not be the most efficient expression system, is nevertheless the one that all regulatory authorities are familiar with and is fully accepted by them. As an evaluator, I do not feel compelled to carefully study all the information about the characteristics of the MCB and WCB if the applicant has used CHO cells for cell banking.  The most important factors are whether the cell banks have been developed in a GMP laboratory, have been characterised according to pharmacopoeial requirements and whether the cells retain genetic stability during the fermentation process.  With plant cells, especially if the applicant has used a transient expression method, there is no cell bank system, which means that there is no common link between different batches. The quality guidelines currently in force would not be applicable to such a method of production.
  • We accept that with CHO cells there will be batch-to-batch variability; however, we also know by now that this variability, for example for potency by ELISA or an in vitro cell based assay, can be defined within specified limits (often 80 to 120%). In the case of a plant-based system, we just don’t know the extent of variation in critical quality attributes or process parameters, i.e., the body of knowledge is currently too limited for an evaluator to perform a risk-based assessment of in-process, drug substance or finished product specifications.
  • Glycosylated biologics are produced as isoforms or structural variants, which although identical in terms of their protein component, differ in the composition of their glycan or carbohydrate chains.  The different variants have the same type of activity, but they may differ in potency and/or pharmacokinetics.  For biosimilars, it is important that the type and number of isomers per batch closely reflect that of innovator batches.  If the same cell expression system is used for the biosimilar as for the reference product, the likelihood will be high that the variants in biosimilar batches will be comparable to those of the reference product. It is doubtful that this would be the case if plants are used as expression system.

Should biopharming be abandoned as an expression system for biologics?  Although, it may not be appropriate for biosimilar production, it does offer an attractive platform for innovator biologics (which require full safety and efficacy data) and especially for subunit vaccines, such as for a recombinant SARS-CoV-2 subunit vaccine.  Since the S-protein (dominant immunogen) of coronaviruses is glycosylated (5),  plant-specific N-glycan epi­topes can be exploited as a target of the host’s immune response and may be beneficial for immune protection, i.e., they may act as adjuvants (6).


  1. Winning with biosimilars: Opportunities in global markets.
  2. Grohs, B.M. et al. (2010). Plant-Produced Trastuzumab Inhibits the Growth of HER2 Positive Cancer Cells. J. Agric. Food Chem, 58, 10056–10063.
  3. Biopharming 101: How plants “grow” medicines.
  4. Jin C. et al (2008). A plant-derived human monoclonal antibody induces an anti-carbohydrate immune response in rabbits. Glycobiology18( 3), 235–241.
  5. Zhao X, Chen H and Wang H (2021). Glycans of SARS-CoV-2 Spike Protein in Virus Infection and Antibody Production. Front. Mol. Biosci. 8: 629873.
  6. Dirk Bosch and Arjen Schots (2010). Plant glycans: Friend or foe in vaccine development? Expert Rev. Vaccines 9(8), 835–842.
0 0 votes
Article Rating
Notify of
Inline Feedbacks
View all comments
Would love your thoughts, please comment.x