1
$\begingroup$

I have recently learned that one of the best ways to isolate a protein in industry is via Chromatography. What are other ways enzymes are isolated in industry? Are there other ways?

$\endgroup$
3
$\begingroup$

Chromatography is one of the best ways to isolate a protein, full stop. It is extremely widely used in both academia and industry.

Before we get into the details, though, a definition is in order. Chromatography is a means of separating components of a mixture using a mobile phase (liquid or gas) traveling though a stationary phase (a solid, gel, or very small beads, frequently with something attached to them). Chemical and/or physical interactions between the components of the dissolved mixture and the stationary phase cause those components to migrate at differential rates, separating them out from each other.

There is a huge variety of protein-stationary phase interactions that can be utilized in chromatography. Microscopic beads made of different substances can be used to create a matrix of pores where larger proteins will migrate more quickly than smaller ones. Epitope tags such as Myc or FLAG can be recombinantly added to a protein's sequence and purified using beads that have epitope-specific monoclonal antibodies covalently bound to them. The tagged protein is "grabbed" by the antibodies while the rest of the mixture washes through, then the binding is reversed by various chemical means that interrupt the protein-antibody interaction, such as high salt concentrations, varied pH levels, or chaotropic agents and the concentrated protein collected all at once.

Sometimes it is known that a certain class of proteins (such as antibodies, for example) will bind very specifically with another protein or protein fragment (in this case, the bacteria-derived Protein A). That protein is covalently bound to beads and used to purify and/or concentrate all antibodies in a solution, regardless of the antibody's specific target. Protein A binds to a section of an antibody known as the Fc region, so that sequence is sometimes fused to a non-antibody recombinantly-expressed protein so that it can be easily purified by this method. Antibodies that are specific for the protein of interest itself can be used for purification.

Many other chemical reactions can be harnessed to cause a protein of interest to bind to a stationary phase substrate and later eluted by reversing or inhibiting the chemical reaction. One example is purification based on electrical charge, called ion-exchange chromatography. If it is known that a protein has an overall positive charge, negatively-charged beads can be used to bind it (and all other positively-charged proteins in the mixture). Once washed, the protein(s) can then be eluted by changing the pH, hence neutralizing their overall charge and breaking the ionic bonds which bound it to the stationary phase. Positively-charged beads can also be used to purify negatively-charged proteins.

There are many other strategies besides the ones I've highlighted here – the term chromatography encompasses a huge range of possibilities for protein purification. I would say that in the majority of large-scale industrial purifications, combinations of different types of chromatography are the only methods used. There are of course other methods involved in the manufacturing of a protein therapeutic, such as cell lysis, filtration, concentration, dialysis, formulation, etc., where chromatography is not used, but for purification it's typically the only one. One reason for this is automation. Chromatography columns can be loaded, washed, eluted, cleaned, and regenerated without the need for any human interaction other than to provide the necessary buffers and materials. It can often also be scaled up quite a bit from benchtop to manufacturing suite, allowing for more material to be processed per batch. And, once everything is tweaked and optimized, you can get extremely pure material out of it.

There are some other protein purification methods, such as organic extraction, density centrifugation, and differential precipitation/solublization, but these are typically (but not always) used more in research and development than manufacturing, for various reasons. You don't want toxic organic chemicals like phenol, chloroform, ether, etc. involved in your manufacturing process if at all possible, because it takes a lot of work to prove to the FDA that all traces have been removed at the end. Centrifugation is not very scaleable, and is rather difficult to automate. One of the issues with precipitation is ensuring that re-solublization is both complete and consistent between batches. The amount of work that goes into validating manufacturing processes for injectable drugs and getting FDA approval is absolutely mind-boggling. Trust me :)

Hopefully I've answered your question! Please let me know if I missed anything or you want to know more about something.

$\endgroup$
7
  • 1
    $\begingroup$ @Andrew Yes, those, experience, and trial and error. If you gave me a new protein to purify, I'd first check journals and look at several protein databases to get as much info as I could. What you're going to be using the protein for is another consideration - are you treating cells with it? Injecting it into mice? Doing biophysical experiments? Do you want just the natural protein, or can I epitope-tag it to ease purification? Another thing I'd check for is whether there are good antibodies available for the protein, as you can use them both for purification itself and for detection. [...] $\endgroup$
    – MattDMo
    Jul 26 '20 at 22:29
  • 1
    $\begingroup$ [...] Just starting out, I'd probably epitope tag it just so I can easily purify it, detect it, and test my protein-specific antibodies on it, then go back to the natural sequence if you need it in that form. You have to be able to look at a sample or column fraction and figure out if your protein of interest is there, and check quantity and purity. Western blotting is usually the best tool for that, along with Coomassie or similar gel staining. Then, it's just a matter of trying different column types based on the protein's predicted biochemical characteristics and seeing what you can get. $\endgroup$
    – MattDMo
    Jul 26 '20 at 22:34
  • 1
    $\begingroup$ You also need to remember that there's not necessarily only one "right" way to do it. I might do ion-exchange followed by nickel followed by size exclusion, whereas someone else might do three different types of ion-exchange. $\endgroup$
    – MattDMo
    Jul 26 '20 at 22:38
  • 1
    $\begingroup$ ”Microscopic agarose beads can be used to create a matrix of pores where smaller proteins will migrate more quickly than larger ones.” Perhaps I am mistaken, but this sounds like you are describing gel filtration. If that’s the case, however, it is larger proteins which elute more quickly than smaller ones. $\endgroup$
    – canadianer
    Jul 27 '20 at 2:11
  • 1
    $\begingroup$ @canadianer good catch. For some reason I was thinking of electrophoresis, where smaller goes faster. $\endgroup$
    – MattDMo
    Jul 27 '20 at 16:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.