“Nowadays, more than one in three approved biopharmaceuticals are glycoproteins, to which sugar chains are attached in genetically modified (mammalian) cell cultures. This is achieved through the post-translational modification of the proteins inside the cell before they are released into the medium,” said Michael Schlüter.

Schlüter has a PhD in chemistry and moved to Biberach in 1985 when Boehringer Ingelheim established Germany’s first biotechnological production site. In 1987, Boehringer produced the first genetically engineered drug, Actilyse®, a tissue plasminogen activator (rt-PA) for the thrombolytic treatment of acute myocardial infarction. The product was co-developed with the American pharmaceutical company Genentech.

“Actilyse is a typical glycoprotein, it has three glycosylation sites and is very heterogeneous,” said Schlüter who was in charge of t-PA analytics at that time. Since 1985, he has been in charge of quality control and developmental analytics of these recombinant drugs.

What is the effect of sugar side chains? They can affect the folding, stability, intracellular protein transport and immunogenicity as well as the functional activity of proteins. These effects were found for EPO, antibodies, interferons and tissue-specific plasminogen activator (t-PA).

What is known about naturally occurring proteins? Sometimes, the structure of the original proteins is known; but the majority of recombinant proteins are at least as well or even better characterised than the originals. This is because recombinant proteins are available in greater quantities and can be accessed more easily.

In Biberach, and also in many other parts of the world, only a few recombinant production cell lines are available. The best known are CHO cells (Chinese hamster ovary) that originate from the ovaries of Chinese hamsters. In addition to mammalian cells, the yeast Saccharomyces cerevisiae is able to glycosylate proteins; Escherichia coli is unable to do so. N-glycosylation is used for biopharmaceuticals, in which oligosaccharides (glycans) are attached to the amino acid asparagine.

The glycosylation of proteins is a highly complex and enzyme-directed site-specific process. It occurs in the endoplasmic reticulum and the Golgi apparatus. The large number of enzymatic and chemical processes involve more than 100 different proteins and genes. The biopharmaceutical industry benefits from a rather fortunate product of evolution; despite this highly complex process, the glycosylation machinery in the three mammalian cells (CHO cell, baby hamster cell or human fibrosarcoma cell) is highly conserved across the species. There are only small differences between human and non-human mammalian glycosylation (Dingermann, p. 92).

Glycosylation depends on a cascade of post-translational modifications of eukaryotic host cells and can vary from cell to cell and even from clone to clone, depending on the fermentation medium, specific productivity and the physiological state of the genetically modified host cells. Something that was true for Actilyse® in 1987, is, according to Schlüter, still valid today: “The process defines the heterogeneity of the product.”

In principal, the process of glycosylation is rather similar. Each glycosylation site leads to a certain range of oligosaccharide patterns. The amino acid sequence determines where sugar chains are attached, which then prompt the cell’s synthesis machinery. The primary structure of the protein determines the glycosylation process. The oligosaccharide attached to the protein depends on the glycosylation site to which the cell transfers a previously homogeneous precursor oligosaccharide, processes it and alters the previously uniform oligosaccharide on its way through the different cell compartments.

For example, the tolerability and efficiency of Actilyse® was improved by shifting the original glycosylation site to a different site. The first-generation drug had a very short half-life of three minutes, and this was due to the high-mannose oligosaccharide attached to the asparagine at position 117.

The second-generation (TNK-tPA) drug has a glycosylation site at position 103 (threonine was replaced by asparagine), the asparagine at position 117 was replaced by glutamine. The altered drug now contains complex-type oligosaccharides. The modifications resulted in a longer plasma half-life (30 minutes). This had advantages for both doctors and patients. It is much easier to administer the drug, for example as a one-time bolus. It does not have to be administered over a period of 90 minutes and can also be applied in emergency ambulances.

Oligosaccharides also have an effect on the specific activity of t-PA. Schlüter names amino acid position 184 as another example that alters the activity of the drug. If no sugar side chain is attached to this position, then the protein has a free asparagine, which leads to a drastic increase in the activity of the drug. If position 184 has a sugar side chain, then the activity drops by up to 55 per cent. These enzymatic differences, which are the direct result of oligosaccharides, can be measured in in vitro assays.

During product development, the oligosaccharides will be analysed (main and minor components), by combining two-dimensional gel electrophoresis, two-dimensional fluid chromatography and mass spectrometry. This enables researchers to look at the glycosylation patterns of individual glycosylation sites and classify them according to the basis of their mass.

Once the oligosaccharides are characterised, it is easier to study their heterogeneity because a typical product standard (material from clinical phase II studies as reference) is used for comparison. This means that the products are compared with an expected value. Once the production process is established, it must be kept constant in order for the product not to change. This also means that the heterogeneity patterns of the sugars attached to the therapeutic proteins have to remain the same. The oligosaccharides of CHO cell culture products are well known because the cells prefer a specific set of oligosaccharides.

At present, the complete glycosylation process proceeds in the cells. “We have to accept the heterogeneity that is generated by a specific production system,” said Michael Schlüter who assumes that the researchers are still far from being able to express specific glyco-forms. However, Schlüter and others regard glycoengineering as a development that is in its infancy and still needs some time before it becomes routine practice. Protein design and glycoengineering will receive greater importance in the future. This will enable biotechnologists to fine-tune therapeutic proteins for therapeutic applications.

Schlüter said that the biotechnological production of antibodies focuses on the development of production cells that no longer produce certain structures and hence block the process at a certain site. He expects that antibodies will, in future, be glycosylated uniformly, but is also sure that glycobiology is still at the developmental stage and remains far from product development. Sugar research has, for a long time, lacked specific methods, but is slowly, somewhat behind schedule, approaching application.

wp - 13 August 2008 © BIOPRO Baden-Württemberg GmbH

References:

Walsh, Gary; Jefferis, Roy: Post-translational modifications in the context of therapeutic proteins, in: Nature Biotechnology 24, 10 (Oktober 2006), p. 1241-1252

Werner, Rolf; Kopp, Kristina; Schlüter, Michael: Glycosylation of therapeutic proteins in different production systems, in: Acta Paediatrica 2007, 96, p. 17-22

Dingermann, Theo: Recombinant therapeutic proteins: Production platforms and challenges, in: Biotechnology Journal 2008, 3, p. 90-97

Walsh, Gary: Biopharmaceutica benchmarks 2006, in: Nature Biotechnology 24, 7 (July 2006), p. 769-776.

Greer, Fiona; Easton, Richard: Glycosylated Bioproducts – Breaking Down the Benefits, in: European Biopharmaceutical Review, Winter 2006, S. 77-83