Suppose I have a his-tagged recombinant protein. If we inject that into people, there is a legitimate concern that the tag would cause adverse immunological responses. Therefore it appears FDA "discourages" the use of such tags, while weighing it against the practicality of purifying untagged proteins and overall benefit. But that's not what my question is about.
It is often said that purification tags should be removed with a site-specific protease. But in practice, almost all such enzymes leave at least one foreign residue. For example TEV protease: The cleavage site is ENLYFQ // G. After removing an N-Terminal tag, we're stuck with that one terminal G. Unless our protein happens to start with a G, this amino acid will be foreign.
My questions are:
- Are there examples of approved therapeutic proteins that had their purification tag removed, but that still contain at least one residual foreign amino acid?
- Have immune responses, or other adverse events, been demonstrated in response to such residual amino acids?
- Have regulatory agencies opined on this issue?
- Anything else you know or believe about this issue?
For the therapeutic proteins I have worked on, the advice I have always gotten is "no tag" (and, if you are expressing in E. coli: no ampicillin!). Very preliminary work can be done with the tagged protein, but you want to re-clone into a tagless system as soon as possible.
A mantra I have been told (from the head of our GMP) is anything you add to the protein at any time has to be accounted for to the FDA. So, if you add a tag, then cleave it off, you have to show that you have fully cleaved the tag from the fusion, and completely purified your protein of interest away from the cleaved tag and from the protease. i.e you will need assays that show the tag and the protease are not present and these assays can add to the expense. Any remaining residues on the protein will need to be shown to not interfere with biological activity (bioequivalence), so you might need the untagged version anyway.
Further, while Ni-NTA is relatively cheap on a lab scale, it gets very expensive when you need multi-liter columns for pilot and production runs.
Overall, it is simpler to not have the tag on at all and to develop a process that allows purification without it.
I am not aware of any added affinity tags on current pharmaceuticals, and would be interested if anyone else does. FDA has required biomimetic therapeutic proteins to hew as closely as possible to the native sequences they are copying, and that usually means no added AAs. Added sequences would also serve to define the material in its patent and this could narrow the protection; as an example hGH from Genentech and Lilly differ only in the presence of the initiating Met and thus allowed two equivalent patents to be granted. A patented protein which uses TEV may become unprotected if someone engineers a TEV which does not leave the nonnative Gly.
Its also the cost of affinity tag purification which discourages their use in commercial products. Since most current therapeutic proteins in development (>250 by some estimates) are some type of antibody, there are many other purification methods available than removable fusion tags and some other methods are often cheaper.
Do we think an extra Gly will directly cause the failure of a promising pharmaceutical? Probably not but so far it has not been worth the risk, or often the added expense, to try this in a commercial protein. This may change in the future, but FDA is a slow moving and conservative governmental agency and it does not (currently) like purification tags.
Hi John, I know some workers who are trying to produce therapeutical proteins by some other means. One of the interesting ways to isolate it from inclusion bodies. the over production of the therapeutical protein is used to tune in a way that maximum of it come in inclusion bodies. Inclusion bodies are mostly (or surely) made up of the recombinant protein. the inclusion bodies get easily separated from the rest of the cells and then some unfolding procedures followed to get the protein in native form. this process neither need tag, nor combursome process of getting read of the tag, nor even primary steps of purification.
For the therapeutic proteins I have worked on, the advice I have always gotten is "no tag" (and, if you are expressing in E. coli: no ampicillin!). Very preliminary work can be done with the tagged protein, but you want to re-clone into a tagless system as soon as possible.
A mantra I have been told (from the head of our GMP) is anything you add to the protein at any time has to be accounted for to the FDA. So, if you add a tag, then cleave it off, you have to show that you have fully cleaved the tag from the fusion, and completely purified your protein of interest away from the cleaved tag and from the protease. i.e you will need assays that show the tag and the protease are not present and these assays can add to the expense. Any remaining residues on the protein will need to be shown to not interfere with biological activity (bioequivalence), so you might need the untagged version anyway.
Further, while Ni-NTA is relatively cheap on a lab scale, it gets very expensive when you need multi-liter columns for pilot and production runs.
Overall, it is simpler to not have the tag on at all and to develop a process that allows purification without it.
Very intersting question. Very rational and prgmatic replies by Edward and Richard.
Hi John, one thing you have also to consider is that at some point you will need GMP-grade Tev protease in your process. How much will depend on its processivity and on how fast you want your process to go. I do not know the current price of such material but last time we check it was not cheap. And also you will have to define some assay to show that you have cleared it from your final product.
- Badges
- Science method