COMMON CHALLENGES IN CONTROL AND TECHNIQUE
BioPharm: What common challenges is the industry facing today when performing glycan analysis?
Rudd: Glycan processing is difficult to control because it involves a complex process that involves the expression of genes (the
genes carry the code for the glycosylating enzymes) as well as the delivery of monosaccharides on nucleotide donors to grow
the glycans. There are many things to control, and nearly 600 proteins are required to build the glycans as well as to transport
the glycoproteins into the correct organelle for complete glycan processing.
If one is over-expressing a protein, it is possible to exhaust some of the glycosylation machinery. It is not uncommon to
find incomplete structures attached to a protein, which indicate that the processes to build the sugars has not operated on
all copies of the glycoprotein going through the secretory pathway. Understanding how to get the cells to express at a level
where the rest of the machinery can cope is just one challenge.
Another difficulty is determining what to do when the glycosylation is going wrong. One needs to understand in depth how media
components and metal ions, for example, can alter glycan analysis. The natural cell is exquisitely tuned and responds very
finely to its environment. It is difficult to replicate this robustness in a bioprocesssor.
BioPharm: You mentioned that industry uses various techniques for glycan analysis. What are the most commonly used instruments and techniques?
Rudd: Glycan analysis depends on a series of separations technologies that exploit different physical properties of the oligosaccharides.
There are several ways to approach it. If you are focused on analyzing released glycans, you need to have an optimized method
for releasing them. There are many glycoconjugates, but from the point of view of the pharmaceutical companies the most commons
ones are N-linked glycans. There is an enzyme called PNGase F that can be used to remove them from proteins of most species.
After that stage, various separations technologies can be used to separate glycans according to mass/charge, charge, size,
and lipophilicity by techniques such as hydrophilic interaction chromatography (HILIC), ion-exchange chromatography, reverse-phase
chromatography, or capillary electrophoresis.
BioPharm: What gaps still exist in glycan-related technology and instrumentation?
Rudd: NIBRT uses a lot of special instrumentation, but we are working to get the analytical technologies miniaturized and as straightforward
as we can. We have a robot that can accommodate 96 well plates, for example, so that one can put the samples on the instrument
and come back later to put the released sugars on a HPLC machine or use mass spectrometry for analysis. We also have the capability
to run linear samples, meaning that we can take samples one at a time from a bioprocesser every few hours. Going forward,
industry will be looking to miniaturization, automation, and, particularly, automated data analysis.
BioPharm: What regulatory expectations exist for glycan analysis when developing a biologic compared with a legacy product?
Rudd: Actually, there is a debate at the moment. Regulators need companies to report critical features of the glycosylation, but
in some cases, it is not clear what "critical" features include. There is a need for more basic research to clarify these
questions. For some molecules, such as IgG, it is known that the Fc glycosylation modulates effector function, so the regulators
can ask for a full glycan analysis of IgG. One can report the sialylated structures, the levels of galactosylated, fucosylated,
and bisected structures, because it is known that each of these features can modulate a function. To get the information,
teams need to perform a complete analysis of IgG and present the data in a way that answers the questions about critical features
that affect the safety and efficacy of their product.
If one is working with erythropoietin, it is necessary to report the percentages of different antennary structures as well
as the extent of sialylation, because this critically affects the pharmacokinetcs of the drug. One needs to report levels
of N-glycol-neuraminic acid, alpha(1,3) linked galactose as well as levels of xylose and alpha(1-3)-linked fucose, because
these are potential antigenic epitopes. In general, regulatory expectations are getting higher because the technology is getting
better. In the past, none of us really understood the implications of glycosylation in therapeutics and regulatory requirements
were to define the glycosylation as well as one could. Our understanding has moved a long way since then, and regulations
are far more demanding.